Pub.
3:5... ..
fiJu-lles
...J.....!....;.. V 5:23.?
, x.
. we. .‘

in. r
2.3." o
I; 5‘. I {lab-’2! 4ilu‘y
1‘. . 4“» .I'. '5 Iv“. .‘I
2 0. 2.5.0... at. a... .95.... m 354!
s... a 2 1?. x. a“ .12.. . mu... .1
sun xi»: Am. i... 9.2.21: 2.“!va r9
ad. a. an»... ....bmun.a nun“...
.. v . .9.

3} \
132.13....
)
Ix

1,

2: ‘ ..
¢ .. w..w,m..mfu:(u ..
H 4
. y ,
, 21, .. .

.33. I... .
rumba? ,
J!
, m . . 9
. . a.
, . ltduzfi ma
. . .3 .. .3. ,
. A .i .
. . Ea)...
. .
. .
j 4 .
.
u)»

. “I m
. . .. . .bm. mu»...

(.1.
3. t.

P. . LE.
5.3:.»
.vahuhn .
‘ .

t
l

"a ; ~ 4.

ma)” ‘ ,_1'

, *gsgéii‘fidséy
i',

ll.

. its.

6 1.. i~:‘
I .5

33$
15‘
3v

2:...

. . .. !. fatal: .3

~ , i. . ‘51:... h.‘

.19»; 9L. :53. 127.91 {Vivallfi

v3 trinsblii! 2.1133}: .
i_ .13.th Is! .03

i?
'i‘v
:.

..
‘
I
3.. a. . 5
t i, a ‘ .
Vuwufla‘in)...‘

l

I
A... a )4 .
19.: l.

1"}. .
vi ‘1 .x‘:
J! r}?

‘I1lnv'i v.5
l. . a
y .. .4 El an! ‘
‘. I a ‘ . 9‘»! III
a ,3). fluff.
.f31...«(u:l .
12.23515 2 A ..
..Ss-4....§v .59.
.finatn’tkbén‘906 it... .
‘tk- 4. "iv l’lil‘
.m‘ a, .2 :...+
v... E. tat-1.;
.. 1.32. i 3 2.1. )y
.2; .95....931nlz
.5... \.
t... .315
ll - n. s ‘
. A1A.U.:I.1S.dit§
. z... .1!!! 1.1 3.1.(r
hum: $3.: . if ....:
a 3:...
I a . 2:; in .
a 7.53:“ .3 is .
3. .i
1 ..fl
“3 ;L. .l
. .~.«»v n1.

 

 

 

"ibfit-S

c1

TY LIBRARIE

flliiili'iiiliiiiiliiiiliifliIHI1hIll|l|||

3 1293 01389 2884

in

 

 

This is to certify that the

dissertation entitled

THE ROLE OF UNKNOWN VOCABULARY AND PRIOR
KNOWLEDGE IN THE COMPREHENSION OF FAMILIAR
AND UNFAMILIAR TEXT WHEN READING IN A

FOREIGN LANGUAGE
presented by

Susana B. Tuero

has been accepted towards fulfillment
of the requirements for A ' i ‘

 

Ph . degree in W

M4224

Major professor

Date 1/4/96

MS U it an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

LIBRARY
Michigan State
Universlty .

 

 

 

PLACE IN RETURN BOX to roman this chockout from your rooord.
TO AVOID FINES rotum on or baton duo duo.

DATE DUE DATE DUE DATE DUE

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Afflrmotlvo Action/Equal Opportunity Inotltutton
Wm:

—‘_,_.___.-——-———

THE ROLE OF UNKNOWN VOCABULARY AND PRIOR KNOWLEDGE IN THE
COMPREHENSION OF FAMILIAR AND UNFAMIUAR TEXT WHEN READING
IN A FOREIGN LANGUAGE

BY

Susana B. Tuero

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of
DOCTOR OF PHILOSOPHY
Department of English

1996

ABSTRACT

THE ROLE OF UNKNOWN VOCABULARY AND PRIOR KNOWLEDGE IN THE
COMPREHENSION OF FAMILIAR AND UNFAMILIAR TEXT WHEN READING
IN A FOREIGN LANGUAGE

BY

Susana B. Tuero

Evidence from studies in the process of reading in the first
language reported strong correlations between word knowledge and
comprehension (Anderson and Freebody, 1981; Nagy, William,
Herman, and Anderson, 1985; Medo and Ryder, 1993). In addition,
some L1 researchers (Flumelhart, 1977; Stanovich, 1980, 1984)
hypothesized that the use of top-down processes (such as prior
knowledge) may compensate for difficulties in bottom-up processes
(such as vocabulary difficulties). Even though studies in L2 reading
have been heavily dependent on L1 reading research, very few L2
studies have investigated this interactive-compensatory hypothesis,
and how vocabulary knowledge relates to comprehension and
information recall in the process of foreign language reading. The
present study was designed to bridge this gap.

Participants in this study were high-school students who had
been studying Spanish for at least tour years at the time the data
were collected. All students read two passages in Spanish, one about
a familiar topic, 'La Noche de Bruias,‘ and the other about an
unfamiliar topic, 'La Diablada.‘ Some of the students read an easy

version, others read the hard version of each passage. The easy

version contained known words and the hard version contained words
that are less frequently used. After reading the passages, the
participants were asked to write a free recall, and then answer
twelve multiple-choice questions to assess their comprehension.
These data were analyzed statistically, and the students written
protocols were individually examined.

Results of this study indicated that a) prior knowledge affected
the students' recalls of supporting units, and their answers to
scriptically implicit questions, b) vocabulary difficulty affected the
students‘ recalls of central and supporting units, and their answers
to textually explicit questions, c) topic familiarity and vocabulary
difficulty did not interact in any of the recall or comprehension
measures, (1) difficult vocabulary appeared to affect the
development of a coherent text in the students' written protocols.

This study concludes that background knowledge and vocabulary
difficulty function independently and affect reading in different
ways. Even though prior knowledge facilitates comprehension,

vocabulary development is vital to foreign language reading.

DEDICATED TO THE MEMORY OF
MY GRANDMA, MARIA CASTANO DE VILAR,
who truly believed that education was

the open door to a better life

ACKNOWLEDGMENTS

This dissertation would not have been possible without the
encouragement and support of several people. I would like to thank
my daughter, Marla Belén, who stood by me during all my years of
study. I would also like to express my appreciation to all the
members of my committee, Dr. Nancy Ainsworth-Vaughn, Dr. Susan M.
Gass, Dr. Marcellette Williams, and Dr. Marilyn Wilson. I am grateful
to Dr. Nancy Alnsworth-Vaughn for her invaluable feedback and her
wise guidance throughout my studies. Dr. Susan M. Gass gave me
advice and support during all my years of study at Michigan State
University. I am grateful to Dr. Marcellette Williams, who always
believed in me, for her invaluable encouragement. I want to thank Dr.
Marilyn Wilson for her help and useful suggestions.

This study owes a great deal to Sabine Helling. Sabine encouraged
and supported me throughout this project. She also helped me with
the materials used in this research and spent long hours reading the
drafts and making useful suggestions. I am also grateful to Gary
Cook Who helped me with the design of this study and the statistical
analysis of the data. I want to thank Dr. Pat Lunn, Gabriela Garbarino,
and Miguel Ballicora for their help during the development of the
materials used in this study.

I also want to express my appreciation to Mr. Mario Torres who
willingly let me come into his classes to collect my data, and to all
the students who participated in this study. Without their help this
research would not have been possible.

V

TABLE OF CONTENTS

List of Tables
List of Figures

Chapter One: Introduction
1.1 Overview of reading research in the first language
1.2 Overview of reading research in a foreign language

Chapter Two: Review of the literature
2.1 Models of reading in the first language
Bottom-up models
Just and Carpenter model
Gough model
LaBerge and Samuels model
Carver model
Top-down models
The psycholinguistic model
Schema theory
Dual coding theory
Conclusions
2.2 Models of reading in a second/foreign language
The psycholinguistic models
Goodman and Smith models
Schema theory
2.3 Vocabulary knowledge and reading comprehension
Conclusions

Chapter Three: Methodology

Hypotheses

3.1 Subjects

3.2 Materials

3.3 Instruments
The written-recall measure
The comprehension measure

3.4 Procedures

v i

viii

11
11
17

20
23
31
35

38
38
39
51
53

56
58
65
69
69
73

.74

Chapter Four: Analysls and Results
4.1 Quantitative analysis
Statistical analysis
Descriptive statistics
Multivariate analysis of variance
4.2 Qualitative analysis
Students' written protocols
'La Noche de Brujas‘
'La Diablada'
4.3 Chapter summary

Chapter Five: Summary and Conclusions
5.1 Summary of research
5.2 Conclusions
5.3 Implications for L2 reading research
5.4 Implications for L2 teaching
Appendix A: Student materials

Appendix B: Additional histograms for recall and
. comprehension

Appendix C: Additional statistical analysis of the data

List of references

vii

81
81
91

100
100

104
105

110
112
121

122

126

144

154

158

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

10:

11:

12:

13:

14:

15:

16:

17:

LIST OF TABLES

:Students' responses not included in the study

: Age and gender of participants

: Sample paragraphs from easy and hard versions
:Sample paragraphs from familiar and unfamiliar text
:Central and supporting units from Halloween

: Central and supporting units from 'La Diablada'

:Schedule of reading conditions

:Descriptive statistics for recall scores

Descriptive statistics for comprehension scores
Effect of background on central and supporting clauses
Effect of vocabulary on central and supporting clauses
Effect of background on comprehension measure
Effect of vocabulary on comprehension measure

Multivariate analysis of covariance for background
knowledge '

Multivariate analysis of covariance for vocabulary

Multivariate analysis of covariance for interaction
between background and vocabulary

viii

59

59

68

7O

72

73

76

:.Options presented to students to rate topic familiarity 78

82

85

86

88

88

89

92

92

93

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

18:

19:

20:

21:

22:

23:

24:

25:

26:

27:.

28:

29:

30:

31:

32:

33:

34:

Multivariate analysis of covariance for background
knowledge

Multivariate analysis of covariance for vocabulary
Multivariate analysis of covariance for interaction
between background and vocabulary

Analysis of variance for recall of central clauses
Analysis of variance for recall of supporting clauses
Analysis of variance for textually explicit questions
Analysis of variance for textually implicit questions

Analysis of variance for scriptically implicit
quesflons

Sample sentences from the students' written recall
Halloween/easy version

Sample sentences from the students' written recall
Halloween/hard version

Sample sentences from students' extra-textual
inferences

Sample sentences from the students' protocols of
the unfamiliar topic

Multivariate analysis of covariance for age
Multivariate analysis of covariance for gender

Multivariate analysis of covariance for years of
language study

Multivariate analysis of covariance for
Spanish-speaking friends/relatives

Multivariate analysis of covariance for students'
visits to Spanish-speaking countries

I X

94
95
95
96
97
98

99
99
100
101
102

105
154

154
154
155

155

Table

Table

Table

Table

Table

35:

36:

37:

38:

39:

Multivariate analysis of covariance for length of stay
in Spanish-speaking countries

Multivariate analysis of covariance for contact .with
the language outside the classroom

Multivariate analysis of covariance for background
knowledge

Multivariate analysis of covariance for vocabulary

Multivariate analysis of covariance for interaction
between background and vocabulary

155

156

156

156

157

Figure
Figure

Figure

Figure
Figure

Figure

Figure

Figure

Figure
Figure
Figure
Figure
Figure
Figure
Figure

Figure

10:

11:

12:

13:

14:

15:

16:

LIST OF FIGURES

: Representation of the 'ship-christening schema'

Dual coding model

Number of years that the students had studied
Spanish

Students who visited Spanish-speaking countries
Spanish-speaking countries visited by the students

Length of students' stay in a Spanish-speaking
country

Students who had/did not have Spanish-speaking
ffiends

Students' language exposure and plans to continue
studying

Recall scores for unfamiliar text

Recall scores for familiar text

Recall scores for hard version

Recall scores for easy version
Comprehension scores for unfamiliar text
Comprehension scores for familiar text
Comprehension scores for hard version

Comprehension scores for easy version
xi

26

32

60

61

62

63

64

65

83

83

84

84

144

144

145

145

Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure

Figure

17:
18:
19:
20:
21:
22:
23:
24:
25:
26:
27:
28:
29:
30:
31:
32:
33:
34:
35:

36:

Central ideas of unfamiliar text

Central ideas of familiar text

Supporting ideas of unfamiliar text
Supporting ideas of familiar text

Central ideas of hard vocabulary

Central ideas of easy vocabulary

Supporting ideas of hard vocabulary
Supporting ideas of easy vocabulary

Explicit questions of unfamiliar text

Explicit questions of familiar text

Implicit questions of unfamiliar text

Implicit questions of familiar text
Scriptically Implicit questions of unfamiliar text
Scriptically implicit questions of familiar text
Explicit questions of hard vocabulary

Explicit questions of easy vocabulary

Implicit questions of hard vocabulary

Implicit questions of easy vocabulary
Scriptically implicit questions of hard vocabulary

Scriptically implicit questions of easy vocabulary

xii

146
146
.147
147
148
148
149
149
150
150
150
151
151
151
152
152
152
153
153

153

CHAPTER ONE
INTRODUCTION

1. 1 OVERVIEW OF READING RESEARCH IN THE FIRST
LANGUAGE

As a result of a renewed interest in cognition during the past
three decades, considerable research has been conducted to examine
and describe what happens in a reader's mind while reading. Some
first-language reading researchers investigated the mechanical
aspects of reading, such as the reader's eye movement (Carpenter &
Just, 1983; Frazier & Rayner, 1982; Just & Carpenter, 1980). Results
of these studies provided interesting information about the words
that the readers fixated on and the length of time the reader fixated
on each word in a text. It was found that more content words
(adjectives, nouns, verbs, and adverbs) than function words
(articles, preposition, and coniuctions) were fixated on. The
proportion of content words that were fixated on was higher even
for skillful readers.

Other researchers hypothesized that the reading process begins
with the reader's recognition of letters, which are then combined to
form words (Gough, 1972). LaBerge and Samuels (1974) moved a step
further and suggested that certain letter combinations are
processed in chunks. Skillful readers become familiar with syllables
that occur frequently in the language and process those syllables as

a single unit.

2
Carver (1977, 1978) claimed that the process of reading begins at

word level. As words are processed, readers comprehend sentences.
Sentence comprehension is an Important mechanism in reading since
the reader is expected to comprehend the writer's thoughts, and the
writer‘s thoughts are expressed in sentences. Carver also
hypothesized that there is an important link between reading and
phonology. He used the term reading to refer to this link which he
considered of major importance in the process of reading.

These reading models that describe the reading process as a
series of stages that occur in a linear fashion, and that start at text
level are usually referred to as text-driven, or bottom-up models. An
important counterweight to bottom-up models was provided by
researchers who viewed the reader as the most important component
of the reading process. Rumelhart and Ortony (1977) brought back
the idea that the readers‘ past experience plays a major role In
reading. It is the readers' knowledge of the world that helps them
bring out the meaning of the text. The readers' prior knowledge is
organized In chunks, usually called schemata , and stored In memory.
It Is during the process of reading that readers use their schemata
to comprehend the text.

Introduced in the late 1960s and early 19798, the ideas presented
by Goodman and Smith had a tremendous impact on reading research.
In their view, the reader Is the most important component in the
process of reading. The reader approaches the’text with hislher
prior knowledge and brings meaning to it. The negotiation process

between the reader and the text results in comprehension.

3
Reading models that consider the reader as the major component

In the reading process are usually referred to as reader-driven, top-
down models. Top-down models have been widely accepted by
teachers, reading specialists, and researchers. Results of a
significant number of studies in first language reading demonstrated
how the readers‘ prior knowledge may help or hinder comprehension
processes (Anderson, Pichert, and Shirey, 1983; Anderson, Reynolds,
Schallert, and Goetz, 1977; Carey, Harste, and Smith, 1981; Pichert
and Anderson, 1977; Shimoda, 1993).

1. 2 OVERVIEW OF READING RESEARCH IN A FOREIGN
LANGUAGE

Top-down cognitive models had a strong impact on research In the
process. of reading In a foreign language. For over a decade, much of
the work done in L2 reading was carried out within the framework of
schema theory (Aron, 1986; Bamitz, 1986; Carrell, 1983a, 1984b;
Carrell and Eisterhold, 1983; Lee, 1986). Results of these studies
demonstrated that not only topic familiarity but also the rhetorical
organization of the text affects reading (Carrell, 1984a, b, 1987;
Kintsch and Greene, 1978). Carrell (1987) claimed that even though
unfamiliar content generally posed more difficulties for the
participants in her study, the rhetorical form of the reading text
was more important In the comprehension of temporal sequences
among the events. Carrell concluded that both content and text form
affect comprehension, but in different ways.

Although all reading researchers recognize the importance of the

readers' prior knowledge in the reading process, the claim that the

4
reader‘s schemata is the major component of the process has

received some criticism recently from researchers and reading
specialists. The critique came from both first- and second-language
reading researchers.

Some first-language reading researchers observed that the term
schema is vague; it has no clear definition (Alba and Hasher, 1983;
Brewer and Treyens, 1981; Paivio, 1986). Sadoski et. al (1991)
wondered, for example, whether “we have an overall schema for
color that Is made up of particular colors (e.g. red, yellow) or
schemata for particular colors that are made up of particular
examples of these colors (e. 9. fire engine red, lemon yellow). (p.
466)

Alba and Hasher (1983), in their review of frequently cited
schema studies, observed that most of these studies used some
obscure and bizarre passages. In addition, they commented about and
Johnson's study (1972),

Their passages contained no explicit, concrete
referents, and without a context to suggest exemplars
for these referents, none Is likely to be inferred.
And they continue,
This should serve to reduce the possibility that one
sentence could cue another at recall. It is not
surprising then that rech of these materials is so
poor; subjects had in effect been presented with a set

of unrelated sentences. (p.220)

5
Even though studies in the process of reading in a second/foreign

language have concentrated almost exclusively on the study of top-
dovm skills, some second-language reading researchers have began
to speculate on the limitations of top-down models. Those studies
have provided important information about the active and
constructive processes involved in reading comprehension, but
reading is a highly complex cognitive process. By concentrating
heavily on the invemigation of only one component, other Important
aspects of the process may be neglected.

Eskey (1988) suggested that top-down models accurately
describe fluent readers for whom the use of higher-level skills is
automatic. Good readers' decoding strategies are also automatic, and
by using appropriate comprehension strategies, they are able to
make predictions and skillfully interpret the text. Thus, top-down
models partially describe the process of reading. Eskey (1988)
clearly expressed his concerns:

but for the less proficient, developing reader-
llte most second language readers- this model does not
provide a picture of the problems such readers must
surmount. (p. 93)

Clarke (1988) recognized the important role that language
proficiency plays in the process of reading. Good first-language
readers may not transfer their adequate reading strategies when
reading In a second language due to linguistic difficulties. It appears
that difficulties at text level, such as vocabulary may "short

circuit" readers' ability to interact with the text.

6
Some first-language reading researchers proposed an interactive

view of the process of reading (Rumelhart, 1977; Stanovich, 1980,
1984). Under this hypothesis, readers confronted with difficulties at
text level (a. 9. because of complex vocabulary) can compensate for
their deficiencies by using higher-level skills, such as their prior
knowledge. Although this interactive-compensatory sounds appealing
in both first- and second-language reading, researchers have failed
to find such interaction in L1 reading (Freebody and Anderson, 1983;
Stahl and Jacobson, 1986).

There is evidence that Indicates that both high level skills and
low level skills affect reading, but they do not interact. Both
vocabulary and topic familiarity were found to affect L1 reading but
in different ways.

In the following chapter, relevant literature will be reviewed.
Models of first-language reading will be described in the first part
of the chapter. Relevant experimental studies will be discussed. The
second part of the chapter presents a review of research studies in

the process of reading in a foreign language.

CHAPTER TWO
REVIEW OF THE LITERATURE

To the observer, reading looks like a simple, effortless skill, a
skill that is expected to be acquired in our childhood and practiced
for the rest of our lives. Like many other skills it is not completely
lost under normal circumstances and results in comprehension.
Contrary to the assumption that it is a simple mechanism, reading,
however, Is a very complex intellectual skill that consists of
various cognitive processes. The use of these cognitive processes

will result In comprehension.

2.1 MODELS OF READING IN THE FIRST LANGUAGE

Bottom-up Reading Models

Substantial research has been done in an attempt to understand
what happens In a reader‘s mind while reading, what factors may
help or hinder the comprehension process. Several cognitive models
have been offered In an attempt to describe the process.

It was during the late 18006 and early 19003 that experimental
psychologists got Interested In the process of reading. In an attempt
to understand reading, substantial research was conducted to study
letters and word recognition, reaction time, and eye fixation. These
were seen as reflections of mental processes and researchers used

their results as a basis for speculation about mental processes.

7

8

Perception and Imagery were also considered important factors in
reading.

However, this Interest in cognition subsided shortly after the
turn of the century. As research focused more and more on
observable behavior, studies on metalinguistlc processes virtually
ceased to exist. The widely accepted behaviorlst theory drew any
interest away from mental processes.

It was not until the 608 that renewed interest in cognition began
to emerge. Once again reading comprehension, attention, information
processing and memory became central in research.

Today research in reading has been enriched by the large number
of professionals coming from different fields. Neurologists,
anthropologists, linguists, and sociologists are trying to describe
the psychological processes that occur in a person's mind while
reading. However, this Is not an easy task. Since those psychological
processes are not open to direct observation, they can only be
inferred by observing what readers do and by relying on what readers
tell us they do while reading.

Several studies that focused on reading analyzed the reader‘s eye
fixation on the words of the text (Carpenter & Just, 1983; Frazier &
Rayner, 1982; Just & Carpenter, 1980). By studying the reader's eye
movement, researchers recorded the words that were fixated on, the
length of time that the reader fixated on each word and which
words were refixated. There are two types of fixation: forward
fixations and regressive flxatlons. Most of a reader's fixations

are forward fixations, from earlier words to later words in the text.

9

It is suggested that if the reader does not encounter major
difficulties, most of the fixations are forward fixations. When
readers have difficulties in understanding the text, they make a
larger number of regressive fixations, fixations to previous parts of
the text.

Just and Carpenter Model. Just and Carpenter (1987) claimed that
by observing the time that a reader spends on various parts of the
text and which words he/she fixates or rereads we can 'have a view‘
of the ongoing cognitive processes. Based on their studies, these
researchers stated that more content words than function words are
fixated and that even skillful readers fixate a high proportion (up to
80 percent) of content words-adjectives, nouns, verbs, and adverbs.
The proportion of function words that are fixated is much smaller.
Readers fixate about 40 percent of the function words-articles,
prepositions and conjunctions (Carpenter and Just, 1983; O'Regan,
1979). These percentages of fixations were made by proficient
readers reading texts on a variety of topics which did not demand a
strong knowledge of the subject.

Although most function words tend to be shorter than content
words, word length does not seem to have been the reason why fewer
function words are fixated. Three-letter content words such as ant
and run were skipped less often than three-letter function words
such as the and and. Just 8: Carpenter (1987) suggest that this may
be the case because the first letter of the function word is encoded
at the same time as the previous content word is processed. Because

of the high frequency occurrence of function words, it is possible

10

the processing of the fixated content word occurs in parallel with
recognition of the following function word. Psycholingulsts such as
Smith (1982), and Goodman (1939) would suggest that the.
explanation lies in the readers' use of linguistic knowledge. The
reader's linguistic knowledge helps him make predictions about the
immediately adjacent word, especially when it is a function word.
Suppose that the reader reads a sentence such as 'It was necessary
to call off the meeting. ' It is quite likely that the function word to
will not be fixated. The grammar knowledge that native speakers of
English have about their language enables them to predict the
occurrence of the word to, making the fixation of this function word
unnecessary.

Although eye fixation research has produced interesting results,
the process of reading cannot be explained only by the results of
this kind of studies. Many models of reading proposed an explanation
of the process beginning at text level. Then, a series of processes
occur in chain, resulting In comprehension. This kind of model that
describes the reading process as a series of stages that occur in a
linear fashion starting at text level Is usually referred to as a
bottom-up reading model. In this view the construction of meaning
begins at letter level; letters are prbcessed and make words which
at the time make phrases, and phrases make sentences. According to
these text-driven models, language is processed in small chunks.
New processed chunks of language are added to the previous ones
until they can be processed at higher mental levels resulting in

comprehension.

11

Gough Model. Some of the most widely known bottom-up models in
reading were proposed in the 1970s In 1972, Gough stated that the
reading process starts the moment readers fixate their eyes on the
text. The reader's eye fixation produces a mental representation.
Gough claimed that readers read letter by letter; as words are
formed, their meaning is processed In a 'mental lexicon' that is
stored in the human brain. When a word makes sense, it is stored in
short-term memory. When the content of short term memory is
understood, comprehension results. Gough does not offer any
explanation about how meaning ls comprehended and stored, nor does
he explain how inferences are made. This view of the comprehension
process as a dictionary of words and meanings is too simplistic.

LaBerge and Samuels Model. Although LaBerge and Samuels' (1974)
model has undergone revision (Samuels and LaBerge, 1983), the
original model ls often cited as another example of a bottom-up
view of reading. LaBerge and Samuels emphasized the role of
attention in the process of comprehension. They assumed that the
reader performs two tasks while reading: decoding and
comprehending. In the case of skilled readers, decoding is an
automatic process. The theory of automaticity holds that as letters
and words become familiar, the reader does not need to pay
attention to decoding and can concentrate on comprehending. For
LaBerge and Samuels, this is the main difference between skillful
readers and poor readers. Skillful readers are able to decode the
printed word automatically and can concentrate their attention on

comprehension. Poor readers, on the other hands are unable to decode

12

automatically. They have to allocate more attention to decoding
resulting in poor comprehension.

LaBerge and Samuels compare Information processing with a
factory. In a factory, raw material enters at one point and passes
through different machines. Each machine has its own purpose, and
the raw material undergoes various changes as it moves along from
one machine to another. Some machines may work faster than others.
When this is the case the raw material is delayed until the next
machine is ready to process the material. At the end a final product
leaves the factory at some other location. LaBerge and Samuels
consider that the same four basic elements that are found in a
factory-raw material, machinery, storage units, and a control
manager-are also found in their Information processing model.

The visual information provided by the text is the raw material
that enters the factory. This visual information consists of letters,
words, figures, pictures and any other kind of visual information
that initiates the process. Once the visual Information has entered
the factory, it is processed by four different machines: visual
memory, phonological memory, episodic memory and semantic
memory.

Visual memory allows the reader to extract information from the
printed marks on the page. Unlike Gough, LaBerge and Samuels
claimed that print is not processed letter by letter. The reader's
visual memory allows him to process not only letters but also
combinations of letters that often form spelling patterns. Proficient

readers, who are familiar with spelling patterns, process groups of

13

letters as a single visual unit. Certain letter combinations that
occur frequently in English such as 'sch,‘ and 'thr,‘ as well as .
prefixes and suffixes, form spelling patterns that are procewed as a
chunk by skillful readers‘.

The second machine in the LaBerge and Samuels model that
processes the incoming visual information is the reader's
phonological memory. This mechanism transforms printed symbols
into sound values. This translation of print into sounds may occur at
different levels such as letter level or whole word level.

According to LaBerge and Samuels, it is in the reader's semantic
memory that word meaning and grammar knowledge is stored.
Semantic memory allows the reader to derive meaning from printed
symbols. It Is also his semantic memory that helps him to make
sense out'of a string of words that conform to the rules of grammar.

For LaBerge and Samuels these three machines (visual,
phonological and semantic memory) are the most important In the
reading process. These researchers hypothesize, however, that
another mechanism, the reader's episodic memory, may affect the
reading process. Episodic memory keeps a record of when, where, and
how a particular piece of infon'natlon was acquired. It is their
episodic memory that hinders young children's comprehension of

texts that are written in an unfamiliar typesetting. Samuels and

 

1 Harbor and Harbor (1981) moved a step further, and hypothesized that readers can

recognize words in a holistic way, without identifying all the constituent letters. This

holistic process occurs as the result of familiarity with spelling patterns, the shape of
the word, and the context in which words occur.

14

LaBerge suggested that knowledge must not be limited to a
particular time, place or setting In order to be maximally useful.

Attention Is one of the most important factors in the LaBerge-
Samuels model. It is called the 'strateglc control manager.‘
Attention “usually refers to the amount of mental energy or effort
required to perform tasks' (Samuels and LaBerge, 1983). These
researchers observe that the amount of mental energy available at a
certain time is limited. It the task requires less mental energy than
the amount that is available, there is not a problem. However, if the
task requires more attention than the amount that is available, the
reader uses a particular reading strategy to overcome his
limitations.

LaBerge and Samuels claim that in order to cope with our energy
limitations, we divide the task into smaller units. Through this
strategy, we are able to perform complex tasks. This process,
however, Is slow and difficult. Sub-units that do not exceed our
attention capacity are processed one at a time. Through practice, the
unskilled person will be able to process the sub-units at a much
lower attention cost. Sub-units will then be processed in groups and
the entire process is sped up.

In this bottom-up view of reading, this concept of speed through
practice can help explain the basic difference between skilled
readers and unskilled readers. In the case of unskilled readers the
different activities that they carry out as they read require much
more mental energy than the amount that is available. Decoding the

words from the text, extracting the meaning of each word,

15

combining the meaning of the different words into phrases and
phrases into sentences that make sense,'and relating the meaning of
the sentences to the rest of the text exceeds the unskilled reader's
attention capacity.

In the case of skilled readers, on the other hand, very little
attention Is used for decoding. Practice helps lessen the amount of
attention for decoding. With practice decoding becomes automatic.
Automatlcity is one of the most important characteristics of skilled
readers. LaBerge and Samuels claim that skilled readers are able to
decode and comprehend simultaneously. These researchers recognize
that word recognition does not guarantee comprehension.
Sometimes, however, the process of word recognition is
successfully accomplished, but it may not result in comprehension.
This may be because of another key component in the model,
attention.

In the LaBerge and Samuels model attention also plays an
important role in the processes of storing and retrieving
information. In order for a reader to store the meaning of a
particular word he has to pay attention. At this stage, however, the
word is stored in what we ordinarily think of as short-term memory.
During this acquisition phase, the reader has to pay attention to how
the word ls pronounced and it Is during this phase that the reader
associates the visual and the phonological representation of the
word. At this early stage, the reader recognizes the word and
retrieves Its meaning but at a high energy cost. With practice,

information about a word will be stored in our long-term memory,

16

and it is at this point that the reader is able to recognize a word and
retrieve it from his/her memory with little attention.

When this model was first introduced In 1974, It suggested a
linear flow of lnforrnatlon from visual memory to semantic memory.
Semantic memory did not have any effect on the other components.
The model has undergone several revisions (Samuels, 1977; Samuels
& LaBerge, 1983) and feedback loops were added to the original
model. These feedback loops show that the components of a system
can interact and affect each other. A well developed subskill can aid
and compensate for other poorly developed subskills in the process.

Although the LaBerge and Samuels model Is often cited as an
example of the bottom-up view of the process of reading, the
addition of feedback loops makes It fall Into a category of
interactive models, which will be discussed later. In this revised
model the feedback loops indicate that all the components are
interconnected. The important result of this interconnection is that
each component can affect any and each of the other components of
the reading process. It later processing stages (semantic memory)
can affect early ones (visual memory), the process can not be
considered truly bottom-up, but rather an Interactive model
(Barnett, 1989).

The LaBerge and Samuels model succeeds in representing, at least
in part, some of the processes that occur while we read. It explains
how important words are in reading. After all, without words, the
process of reading would never occur. LaBerge and Samuels do

acknowledge, however that reading is a complex process and that the

17

ability to decode words does not guarantee comprehension. This lack
of comprehension may be explained through the automaticity theory.
The theory suggests that, for unskilled readers, comprehension is
hindered because decoding uses too much attention. A similar
problem has been found with college students. Samuels and LaBerge
(1983) point out that this occurs because the students use very
little attention in decoding. Thelr attention is freed and directed
elsewhere. The students perceive decoding as a non-demanding
process and instead of focusing their attention on “understanding
and recalling the author's viewpoints,“ they direct their attention
elsewhere, sometimes to matters totally unrelated to the text they
are reading.

The LaBerge-Samuels model does not elucidate what other skills
are necessary to achieve comprehension. Attention is important in
reading, but since It is not the only mechanism needed in reading,
what other skills do efficient readers use? Perhaps, as Otto (1983)
suggests, reading models offer a partial view of the reading process.

Carver Model. Carver (1977-78) hypothesized that the process of
reading begins at word level. For Carver the sentence is the unitary
expression of thought and, since the main purpose of reading Is to
comprehend the thoughts of a writer, sentence comprehension is an
important mechanism In reading. Carver uses the term rauding to
refer to what he considers a major process in reading. The term
rauding refers to the link between reading and phonology. Carver
considers this link of major importance. During the reading process,

the reader says each word lntemally and this internal articulation

18

helps in the process of comprehension. Carver's model is often cited
as a bottom-up view of reading because during rauding, each word
that makes up a sentence is checked to determine whether it
matches the complete thought that is being introduced.

Barnett (1989) pointed out that Carver's model of reading does
not include some reading skills arch as skimming and scanning. If
during the process of reading, the reader checks each word for
comprehension, the process must be different when the reader is
going over the text only to get the gist of it or when he is looking
for specific information and not checking every single word in the
text.

However, Carver (1981) did consider skimming and scanning as he
claimed that these are different types of reading processes. In fact
he claimed that it Is misleading to talk about “the reading process"
As we read, we take different steps that lead to different results.
Carver (1990) stated that there are five types of reading processes:
(1) scanning; (2) skimming; (3) rauding; (4) studying/teaming; and
(5) memorizing. These types of processes are directly associated
with reading rate -that Is, how fast words are covered-and final
results while reading. Carver proposed that these five basic
processes be considered as different “gears“ that operate at
different rates and at different levels of power, from first gear to
fifth gear. Scanning and skimming are the higher or fastest gears
-fifth and fourth gears respectively. When scanning, readers are
searching for a specific word or words. This process involves only

lexical accessing in order to avoid missing the target word.

19

Skimmlng involves not only lexical access but also semantic
encoding. In the skimming process the reader has to attach a
semantic encoding component to each word that is accessed. Carver
claimed that the skimming process is more powerful than the
scanning process because It involves two components -Iexical
access and semantic encoding-instead of only one as in scanning
(lexical accessing).

Rauding, the third gear, is considered the most important process.
It Is the type of reading process that most of us use in normal,
natural reading situations. According to Carver, rauding is the type
of process that results in cornprehenslon. The reader recognizes each
consecutive word and Integrates the words into a complete thought.
His position has not changed since his earliest research and he still
believes that sentence comprehension Is the most important step In
reading.

The teaming and memorizing processes, the second and first
gears respectively, are slower and more powerful than scanning,
skimming, and rauding. During the learning process, the reader's goal
is to understand and remember the Ideas presented In the text.
During this process, the reader continuously checks his ideas to
determine whether he will remember them later. As the idea-
remembering component ls added, the reading process becomes more
powerful and takes additional time.

The memorizing process, the first gear, is the slowest and most
powerful gear. In this process, the reader's goal is to recall specific

facts later. This goal is achieved by adding another component to the

20

reading process, called fact rehearsing. This component involves a
great deal of regressing and repeating information so as to make
sure that thoughts and ideas will be successfully recalled'later.

Carver's reading model is a current view of reading as a bottom-
up process. Reading begins at word level. The meaning of each word
is integrated into the meaning of the 'new‘ sentence that results in
the thought represented by the sentence. This 'new' thought is not
isolated but ls related to the thoughts that rearlted from the
previous sentences.

Although Carver's view of reading as different types of processes
rather than a single cognitive process seems sensible, this model
provides a partial view of the reading process. Carver does not
acknowledge the important role that the reader plays in the process.

He does not discuss how inferences are made.

Top-down Models

The Psycholinguistr‘c Models. The theories presented by Goodman
and Smith have constituted an important counterweight to bottom-
up or text-driven models of reading. Goodman (1967) refers to the
process of reading as “a psycholinguistic game." According to the
Goodman's model the reader approaches the text with his prior
experience, with his knowledge of the world. It is the reader's
knowledge that makes him less dependent on the text. The reader is
central in this model. It is the reader who brings meaning to the
text. Goodman views reading as a process of negotiation between the

reader and the text. Comprehension occurs as the result of this

21

negotiation process. This model identifies four mechanisms that
readers use while reading: predicting, sampling, confirming and
correcting. The reader's knowledge of the language helps him make
predictions about the grammatical stnrcture of the text. At the same
time the reader is also making semantic predictions. These
syntactic and semantic predictions are going to be tested during the
sampling process. It is during sampling that the reader confirms or
corrects his predictions. Goodman claims that the reader does not
need to process every letter or word in the text to confirm his
prediction. It is during this stage that the reader has to make
adjustments if the print does not match his semantic predictions or
if the grammatical structure predicted is not there.

Goodman and his followers carried out several studies to
demonstrate how readers use their knowledge to make semantic and
grammatical predictions. They used a technique that they called
missus analysis . Subjects were asked to read aloud texts that they
had not seen before. Every instance in which the reader departed
from the text was considered a miscue. It was believed that readers
use three language cue systems when reading: the syntactic cue
system, the semantic cue system, and the grapho-phonemic cue
system. Miscues were later analyzed on the assumption that the
readers' departures' from the text reflected the ongoing cognitive
process while constructing meaning through predicting, sampling,
confirming and correcting (Goodman, 1981). The analysis of the
reader's miscues indicates which cue system the reader used or did

not use. Results from all these studies indicate that skillful readers

22

miscue within the same domain and do not correct themselves as
long as there is no interference with meaning. Good readers' miscues
are grammatically correct and semantically appropriate. Readers
generally miscue using at least two cue systems; the following are
examples of miscues provided by Weaver (1988) in her book R_eag_'r_tg

Prgess and Practice (p. 4):

truck
'The little monkey had it.‘ (syntactic miscue)

in the water
'... swimming and playing water games. (syntactic and semantic
miscues)

lmprentice
school was not as important for girls as...‘ (grapho-phonemic
miscue)

Like Goodman, Smith considered the reader the most important
component in the process of reading. Reading is an active process
during which the reader brings meaning to the text. (Smith, 1971)
Smith observed four basic characteristics in reading: It is
purposeful, selective, anticipatory and It is based on prior
knowledge. Reading is purposeful because readers have specific
reasons or goals when they read; it is selective because readers
consider what is relevant to their goals; It is based on
comprehension because readers bring their knowledge to reading.
Their prior knowledge, what Smith calls non-visual information, Is
what helps readers bring meaning to print. As they read, new
information is added to the reader's prior knowledge. Reading is

anticipatory because the reader's prior knowledge, his goals in

23

reading, and his expectation of comprehension help him/her
anticipate the context of the text.

Models of reading that are reader-driven are uarally referred as
top-down models. Like bottom-up models, top-down models
describe the reading process as a linear process. Unlike bottom-up
models, the process of reading starts at high-level mental stages
and moves down to the text.

Schema Theory. Schema theory is the cognitive model that has had
the strongest impact on reading models. Its origin can be traced back
to the beginning of this century. Schema theory was first introduced
as a theory to explain how knOWIedge is stored in memory. Sir
Francis Bartlett Is usually cited as the first psychologist to use the
term schema. However, Anderson and Pearson (1989) observe that
the idea that knowledge may be represented schematically was
introduced by the Gestalt psychologists, and Sadoski, Paivlo, and
Goetz (1991) claim that the idea can be traced to Kant's Critique of
Pure Reason.

Rumelhart and Ortony (1977) are often cited as the first
researchers In cognitive psychology to bring back the idea that
abstract knowledge is stored In cognitive schemas. A schema Is a
'chunk' of knowledge that is stored in memory. Our experience, our
knowledge of the world Is stored in cognitive schemas, and all new
experience is perceived and analyzed in the light of our prior
knowledge. In reading, it is the readers' schemas that play an
important role in the process. Rumelhart and Ortony also recognize

a place for the text In the reading process. The syntactic, semantic,

24

lexical, and orthographic information that a reader has available Is
going to Influence his/her perception of print. The reader's prior
knowledge interacts with his/her semantic and syntactic knowledge
in the reading process.

Models that describe the reading process as an interactive
process between the reader and the written text are usually referred
as interactive reading models. As in bottom-up models, the print is
Important; but unlike bottom-up models the reader's mental
activities are thought to have an impact on comprehension. Reading
is not a linear process but rather a process in which all Its
components are interconnected and can affect each other.

During nearly two decades an impressive number of studies have
been carried out within the framework of schema theory in reading
research. Different terms have been used to refer to how abstract
knowledge is stored in memory. Minsky (1975) talks about frames to
refer to the representation of knowledge in memory. Kintsch & van
Dijk (1978) use the term macrostructures while Schank (1982)
talks of plans to refer to the cognitive structures stored in memory.
However the terms schema and schemata or schemes have gained
preference and most researchers use the term schema theory to
refer to all models that consider prior knowtedge to be a crucial
point in comprehension.

According to Anderson and Pearson (1989) “a schema Is an
abstract knowledge structure" (p. 42). Our everyday experience
results In knowledge that is abstracted and retained in memory. In

the jargon of scheme theory, the parts that constitute a schema are

25

sometimes called 'nodes,‘ 'slots,‘ or 'frameworks.' According to this
theory an event or situation activates a particular schema that is
used to interpret the situation. That means that the new situation Is
going to be interpreted in the light of the activated schema.
Anderson and Pearson believe that once a schema Is activated the
nodes that are part of that particular schema are likely to be
instantiated. On the other hand, an Instantiated node Is not likely to
activate a particular schema. As an example, Anderson and Pearson
refer to 'the Ship Christening' schema. A person's Ship Christening
schema is likely to remind him of components of the schema such as
'new ship,’ 'bottle broken on bow,‘ and 'celebrity.‘ But the mention of
a celebrity is not likely to activate the Ship Christening schema. It
is asarmed that the same node Is shared by different cognitive
structures and that some of them are more salient than other In a
particular schema.

Based on Anderson and Pearson's explanation, we can understand
that there is a categorization of the components that make up a
schema. However, how this categorization Is achieved is not
discussed. Why is it that some components are salient while others
are not? It may be reasonable to claim that a person ls more
familiar with the concept of 'broken bottle' in situations other than
the christening of a ship. It Is quite likely that most of us have
acquired the concept of 'broken bottle' in different situations, and
only a few from christening a ship.

As can be seen in figure 1, In the schema diagram provided by

Anderson and Pearson, the concept 'broken bottle' is diagrammed as a

26

node In the Ship Christening schema. However, the idea of what
exactly constitutes a schema and what exactly a node Is, does not
seem to be well understood. Can a component of a schema act at a
certain point as a schema Itself? Can the concept of 'broken bottle'
which is considered a node In the ship christening schema be

considered a schema?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Done 131'
Celebrity
To Bless Infill/es
Sh'p Ship
Ship-
Christening
Schema
. Bottle
In Dry /
Broken
Dock l. on Bow
Done
Just Before
Launching
Suspended Contains

from Rope Champagne

 

 

 

 

 

 

Figure 1. Representation of the 'ship-christening sohema.‘ From A sdvema-
meoretic view of basic processes, (p. 43) by R. C. Anderson and P. D. Pearson,
1988.

Although it is an accepted fact that readers use their background

knowledge to construct meaning, some psychologists claim that the

27

form schema has left us with the false Idea that we understand
high-level stages in reading much better than we actually do.
Sadoski et al. (1991) commented: I

the pervasive use of the term smema has at times created an Illusion

of consensus and has left the impression that we have a more profomd

understanding of cognition in reading than we do. (page 465)

The term schema has become synonymous with prior knowledge
or background information. However, how knowledge is actually
abstracted from each experience, how much knowledge Is
abstracted, and how abstraction occurs Is still a puzzle In cognitive
psychology (Sadoski et al., 1991).

Alba and Hasher (1983) Identified five central processes in
schema theory. Four of these processes are identified as encoding
processes and one as a retrieval process. The four encoding
processes are selection, abstraction, Interpretation, and integration.
Through selection, some information is selected for representation.
Abstraction Is the process that stores meaning but not the
syntactical and lexical structure. Through the interpretation
process, relevant background knowledge Is Instantiated to facilitate
comprehension. As a result of the previous processes a
representation is formed in memory. Through the retrieval process,
called reconstnrction, we rebuild an episode or event. Alba and
Hasher concluded that memory for complex events Is richer and more
detailed that these processes would allow for. Besides, the process

of Integration does not appear to be essential. Research shows that

28

unintegrated Information is sometimes available for recall (Alba and
Hasher, 1983; Sadoski et al., 1991).

Despite the fact that several critiques of schema theory have
been published (lran-Nejad, 1987b; Paivlo, 1986; Sadoski et al.,
1991), Its Influence on reading in first and second language is still
very strong.

During the past decade, several studies have been carried out to
demonstrate how schematic knowledge help readers select,
comprehend and retrieve information. Among the most Influential
schema studies often referred to are those carried out by Pichert
and Anderson (1977); Anderson and Pichert (1978); Anderson,
Pichert and Shirey (1983). In these studies, the researchers had
their subjects read a passage involving two boys who visited a house
and noticed the existence of different things In the house arch as,
bicycles, expensive china, and a rare coin collection. At the same
time there is reference to some characteristics of the house Itself
such as the existence of a new fireplace, a leaky roof, etc. The
researchers divided their subjects into three groups. Two groups
were asked to assume the perspective of two different persona. The
arbjects In one group were prospective home buyers while the
subjects In the other groups were told to read the passage from the
perspective of a thief. The third group was a control group with no
assigned perspective.

Results from these studies Indicate that the subjects recalled
more information that was relevant from the perspective of the

persona they had been assigned. Home buyers remembered things that

29

might affect the house as real estate while burglars recalled
valuable objects that could be removed. When the subjects were ‘
Instructed to change perspectives for a second recall (without
rereading the passage) they were able to remember facts that were
important from the second perspective but that they had not
previously recalled since those facts were unimportant from the
first perspective. These results support Oakhill and Garnham's
claims that in addition to what happens when the text ls read, what
happens at the time of the recall affects what Is remembered
(Oakhill and Gamham, 1988).

Shimoda (1993) had his arbjects read two short excerpts from
psychology textbooks and two from civil engineering textbooks. The
participants In this study were senior-level college students whose
majors we're psychology and civil engineering. The rearlts of this
study support schema theory. Subjects reading familiar topics were
able to read faster and attain better comprehension.

Several studies have looked at how prior knowledge and topic
interest affect reading comprehension (Asher, Hymel, & Wigfield,
1978; Bellonl 8r Jongsma, 1978; Stevens, 1980). Results from these
studies Indicate that reading comprehension Is enhanced when
arbjects read material that was interesting to them. Baldwin,
Peleg-Bruckner, and McClIntock (1985) also studied the effect of
higher knowledge and topic interest on reading comprehension and
added a new variable, sex differences. Their 41 seventh- and
eighth-grade students completed a 10-item interest Inventory and

took a 100-item prior knowledge test. The subjects were considered

30

fluent readers based on their scores on the California Achievement
Test. Although results from these study suggest that both prior ’
knowledge and topic interest affect reading comprehension, these
two factors seem to be autonomous. Baldwin et al. concluded that
the notion that prior knowledge and topic interest are highly
correlated may be true for adult readers. As people get older they
tend to specialize In areas that they find interesting. Therefore,
knowledge and interest will correspond closely. The situation can be
quite different for school children who are forced to study topics
that may not be Interesting for them. The result would be readers
who have high knowledge of topics that are not appealing to them.
Baldwin et al. observed that there was some Interaction between
topic Interest and gender. These researchers found that topic
Interest had a greater effect on boys than on girls.

Although the effect of prior knowledge on reading has been
extensively studied, how much it affects reading and how much it
facilitates comprehension still remains unclear. Some researchers
have found that prior knowledge may sometimes Interfere with
reading comprehenaon. Like Peeck, van den Bosch, and Kreupeting
(1982), Alverrnann, Smith, and Readance (1985) found that
activation of prior knowledge did not facilitate comprehension. In
fact, results indicate that readers whose prior knowledge conflicted
With the Information they read did not 'negotiate' the new
Information and 're-accommodate' their background knowledge.
Instead they permitted their prior knowledge to override the new

information. Although this did not seem to have much effect on the

31

multiple choice test measure, it did have a detrimental effect on the
recall measures. ,

Alvermann et al. (1985) also raise the issue of the Importance of
the type of measure that is used to assess reading comprehension.
Recall has become a common meaarre for comprehension assessment
In L1 reading. Johnston (1983) pointed out that recall is the “most
straightforward assessment“ of the Interaction between the reader
and the text. More recently, Hayes (1989) claimed that recall
analysis Is the most powerful tool that cognitive psychology has to
describe psychological processes. Although multiple-choice tests
are also a common tool to collect data, results from studies that
used only this type of comprehenaon measure should be evaluated
cautiously. Benhardt (1991) Indicates that a reader's understanding
of the text may be Influenced by questions or other type of
information that he has to read as his reading comprehension Is
assessed.

Dual Coding Theory. Dual coding theory Is a theory of cognition
that has regained Interest among psychologists and reading
researchers. Although dual coding theory Is contemporary to schema
theory (it was Introduced in the early 19709) it has not attracted as
much research as schema theory. Sadosky et al. (1991) claim that
the main reason why dual coding theory remained almost ignored all
these years is because of Its imagery component. It was not until
recently that imagery began to be considered Important in the

process of storing information In memory.

32

Like schema theory, dual coding theory is a theory of cognition. It
attempts to explain how knowledge is abstracted and stored In
memory, how linguistic and non-linguistic representations interact
In cognition. Supporters of dual coding theory believe that cognition
consists of two mental subsystems that specialize In the
representation and pocessing of Information. These two
subsystems, verbal and non-verbal, have the capacity to operate in
an integrated manner. This Integration facilitates the processing of

Information In reading.

VERBAL STIMULI NONVERBAL STIMULI

I

REPRESEN TATIONAL CONNECTIONS

WWW

 

 

SENSORY SYSTEMS

 

 

 

 

 

DOSSV

FEF'ERENTIAL
CW5

 

 

 

 

ASSOCIATNE STRUCTURE

Zm4m<m r>mmm<

 

 

WLWLS SAILVI

gm4m<m r>mmm<zoz

 

VERBAL RESPONSES NONVERBAL RESPONSES

Figure 2. Dual coding model. From Mental representations: A dual ceding approach
(p. 67) by A. Palvio, 1986. New York: Oxford University Press.

Figure 2 presents a model of the two subsystems, and the three

components assumed by dual coding theory. Information Is

33

continually detected by our sensory systems. This Information
presented as verbal and non-verbal stimuli activates mental _
representations. These mental representations are called logogens In
the verbal arbsystem and imagens In the nonverbal arbsystem. The
verbal subsystem specializes in language. It involves the processing
of linguistic stimuli. The non-verbal systems Involves the
generation of mental Images while processing visual, auditory and
affective information.

Sadoski et at. (1991) claim that a Iogogen Is a representational
unit that could correspond to a grapheme, a word, or to a larger
familiar unlt arch as a phrase or a sentence. Representations in the
nonverbal subsystem are “organized more in the form of holistic
nosed sets,“ e.g. eyes, noses, and mouths could be perceived
separately, but are normally perceived as parts of a face.
Representational units within a subsystem display associative
characteristics labeled associative structure In Figure 2. This
associative structure among units occurs as the result of experience
or teaming.

The model also contains three major components (Mayer and
Anderson, 1991, 1992). Components 1 and 2 Involve building
representational connections between the presented lnforrnatlon and
the subsystems. Component 3 Involves building referential
connections between the two arbsystems. These Inter-system
relations account for the great flexibility observed in human

cognition. It Is through these referential connections that language

34

can evoke some imagery or no Imagery at all, and that mental Images
can boost language. _
Dual coding theory is not inconsistent with schema theory, but it
seems to provide a more complete and better articulated model of
cognition. Some results In reading research can be better explained
by dual coding theory than by schema theory. According to schema
theory mental representations are stored in an abstract form, a
schema. A schema selects the presented information, Interprets It,
and stores It in an Integrated form. It is also a schema that governs
the retrieval of information. If this Is the case, readers' protocols
should be quite consistent whether the task is to report on what
they read or report on the images they had in their mind \M'IIIB they
read. However, results from some recent studies (Long, Wrnograd, &
Bridge, 1989; Sadoski, 1983, 1985; Sadoski, Goetz, Olivares, Lee, &
Roberts, 1990) have shown that lnforrnation in reports about
imagery overlaps only partially with information In verbal reports.
According to Sadoski, Paivio, 8r Goetz (1991) this apparent
inconsistency can be easily Interpreted through dual coding theory.
It incoming Information is stored under two mental representations
In two separate subsystems, some variations can occur at the time
of the retrieval. Supporters of dual coding theory claim that
although these two arbsystems have the capacity to operate In an
integrated manner, they can also operate separately. Some questions
still remain. Future studies will have to investigate, for instance,
what causes the two arbsystems to operate separately, and the

circumstances under which this phenomenon occurs.

35

Support to dual coding theory has come from recent research that
has focused on the study of the important role that imagery plays in
Information recall, and the process of reading (Danan, 1992; Intraub
and Hoffman, 1992; Mayor 8 Anderson, 1991, 1992; Gambrell &
Jawltz, 1993; Kulhavy, et at, 1993). Mental Images can be built by
the reader during verbal elaborations (Stein, Brock, Ballard, & Vye,
1987), and comprehension can be facilitated by presenting
information visually, that is by presenting pictures which Illustrate
the text.

Mayer and Anderson's findings (1991, 1992) are consistent with a
dual coding model. When reading scientific texts, subjects who used
explanative pictures were able to understand better than those who
did not have pictures. Because the material was encoded vlarally and
verbally in an integrated manner, the subjects were able to
remember and transfer information more easily than when the

material was presented using one mode, verbally or visually.

Conclusions

For the past two decades, there has been a renewed Interest in
cognition. Much research has been devoted to Identifying and
explaining the cognitive mechanisms that are involved In the reading
process. Despite the effort, full agreement on how comprehension Is
attained has not yet been reached.

When we think of the many variables that affect reading and
reading comprehension ~type of text, syntax, vocabulary, age of the

reader, his motivation, Interest, purpose- It becomes easier to

36

understand why it Is hard to come to general agreement. in addition,
the type of research that is done is usually affected by the
scientific philosophy of the moment. .

The many competing views of the reading process have led to the
development of different theories. These competing views, however,
always consider some basic components on which they choose to
either agree or disagree. These areas of agreement or disagreement
Include the role of the text, the role of the reader and the reader's
background knowledge, and the purpose of the reader.

The role of the text. Some researchers view reading as the act of
extracting meaning encoded in the text by using different language
systems such as orthographic structures, lexicon, and syntax. At the
other end of the continuum are the researchers who believe that the
text is just a blueprint that has no meaning in Itself. Only the reader
can bring meaning to a written text. Meaning is always relative.
Since It only exists in the heads of the readers, a text has the
potential of generating different meanings.

The role of the reader. Some theories perceive the reader as a
passive recipient of information. Through the use of different
language systems, the author 'pours' meaning on the text, and a
skillful reader Is the one who is able to successfully obtain the
meaning that the author encoded on the paper. Other theories see the
reader as the key component in the reading process. No text has
meaning until the reader brings meaning to it.

The reader's background knowledge. This element is closely

related to how the previous two components are perceived.

37

Researchers who claim that the process of reading starts at the text
level, that the text is the key component, and that the reader's job is
to attempt to retrieve the message encoded by the writer may
totally ignore the reader‘s background knowledge when discussing
the reading process. At the opposite end are the theorists who
suggest that meaning is always constructed in the light of the
reader's background knowtedge. It Is the reader's prior knowledge
that helps him create meaning.

The purpose of the reader. Theories that hold that the text has a
precise meaning do not recognize that the reader's purpose Is of any
importance in the reading process. When the written text is placed
at the center of the process, the purpose of the reader is viewed as
Irrelevant in reading. Considering the purpose of the reader
irrelevant seems to be one of the most difficult positions to sustain.
Common knowledge and personal experience tell us that the way we
read a book we chose to read is different from the way we read a
scientific text on which we are going to be tested. Other theories
view the reader‘s purpose as having a significant impact on how
meaning is constructed. In fact, these theorists suggest that the
reader starts with a purpose and predicts meaning before he looks at
the text.

The perspective adopted here is what some theorists have called
transactive theories (Caimey, 1990). Meaning is not in the text, nor
is it In the reader. Both components are Important in the process of
reading. On one hand, the process of reading would not exist without

text. The text is important as it has been shown by many research

38

studies; many of the text characteristics are of significant
importance in the process of reading. Other research studies have
investigated the relationship between. the reader‘s linguistic
knowledge and reading comprehension. Results indicate, for
instance, that phonological information affects comprehension
processes (McCutchen & Perfetti, 1982; McCutchen, Bell, France &
Perfetti, 1991). Research has also demonstrated that the reader is a
very important component of the process. it is the reader's prior
experience, his knowledge of the world that will have a strong
effect on comprehension. Comprehension results from the
transaction between the reader and the text. As Rosenbiatt suggests
the reader creates a 'poem.’ This poem is different from the text on

the page and from the text in the reader's head.

2.2 MODELS OF READING IN SECOND/FOREIGN LANGUAGE

The Psycholingulstlc Model

Goodman and Smith Model. Empirical studies in the process of
reading in a second or foreign language have been heavily dependent
on the first language reading research. An examination of the
literature about the process of reading in a secOnd/foreign language
provides rather surprising findings. A close review of the literature
reveals that recent research In the field has been dominated by two
L1 theoretical frameworks: the Goodman's and Smith's
psycholinguistic model, and by schema theory. Bernhardt (1991)

examined 25 journals that “are of greatest interest to and accessed

39

by teachers and researchers in the area of second language learning
in the United States, Canada, and English-speaking Europe and Asia.“
(p. 21) Her investigation revealed that from 1974 to 1988,. 66.4
percent of the L1 citations in L2 reading studies referred either to
Goodman or to Smith. There seems to be no clear reason to explain
such a dominance of the psycholinguistic model. Bernhardt (1991)
claims that it was due to a "lack of awareness and perception of the
capabilities of models other than those of Goodman and Smith to
explain the second language reading phenomena." (p. 22)

Schema Theory. For over a decade, the schema theory of cognition
has also had a very strong impact on ESL reading. Much of the work In
second language reading research has been carried out within the
frame of schema theory (Alderson and Urquhart, 1988; Carrell,
1983a, b, "1984b; Carrell and Eisterhold, 1983; Lee, 1986). Extensive
research was done on how prior knowledge affects L2 reading. This
is hardly surprising, since a large number of L1 reading studies have
revealed the importance of background knowledge in reading
comprehension.

Basically, background knowledge studies can be divided into three
broad categories. First, some studies examined how 'cultural
schemata' affect reading. Second, other L2 reading researchers
studied how 'topic familiarity' influences the process of reading.
Third, there are studies that have looked into the readers' linguistic
proficiency, and text features such as syntax and vocabulary as
variables affecting the process of reading in a second language. In

some cases two variables were taken Into consideration, such as

40

topic familiarity and language proficiency (Mohammed and Swales,
1984; Alderson and Urquhart, 1988) and cultural origin and
syntactic complexity of the text (Johnson, 1981). ’

Johnson (1982) observed that prior knowledge influenced her ESL
readers' comprehension when they were presented with a passage
that referred to Halloween. Johnson hypothesized that topic
familiarity affected comprehension. More recently Pritchard (1990)
determined that the reader's cultural schemata and the cultural
perspectives of the text influenced not only reading comprehension
but also reading strategies. in this study, 60 11th-grade students
(30 American and 30 Palauan) read two letters. Each letter
described the events surrounding a typical funeral In one of the two
cultures. A taxonomy of processing strategies was developed based
on the subjects protocols. Five categories were identified: a)
developing awareness, such as recognizing loss of concentration or
failure to understand part of the text; b) accepting ambiguity: for
example, formulating a question; c) establishing intrasentential
ties: for instance, rereading and paraphrasing; d) establishing
intersentential ties, relating a sentence to a previous portion of the
text, and 9) using background knowledge: for example, referring to a
previous passage and speculating beyond the information presented
in the text. Results from this study indicate that strategies used by
the readers differed according to whether or not they were reading a
culturally familiar or unfamiliar passage. When reading a culturally
familiar passage, subjects used more strategies that were from

categories 'd' and 'e'. Pritchard claims that by establishing

41

intersentential ties and using their prior knowledge, his subjects
were able to 'develop a unified meaning of the text.‘ .

Prior knowledge also plays a very important role In lnferencing, a
cognitive process used to construct meaning. lnferencing in reading
refers to the mechanism through which the reader gets information
beyond what is explicit in the text. Consider the following example
provided by Rumelhart (1977b):

Mary heard the ice-cream man coming down
the street. She remembered her birthday

money and rushed into the house.

This sentence can be easily understood by readers who know
about ice-cream and ice-cream trucks. For some readers it is easy
to 'infer' that Mary likes ice-cream and that she wants to get some
with the money she received for her birthday. This sentence does not
say that Mary likes ice-cream; nor does it say why she ran into the
house. We are able to infer, however, that Mary ran into the house to
get money to buy ice-cream. it is our prior knowledge that helps us
get this Information. in some cultures, on the other hand, ice-cream
can only be purchased at special places, such as ice-cream parlors.
Someone who comes from such a social group will have difficulty in
understanding what an ice-cream truck is and how Alice could have
'identified it' by 'hearing' it. Why did she run into the house? Was she
scared by the man? Did she run into the house looking for a shelter?

Or did she think that her money might be stolen? Consequently, it

42

may be claimed that lack of relevant background information can
hinder reading comprehension. '

Other studies in L2 reading have examined the effect of topic
familiarity and text structure on reading comprehension (Alderson
and Urquhart, 1988; Carrell, 1984; Carrell, 1987; Nunan, 1985).
Results from these studies have provided evidence that familiarity
with text content and structure facilitate comprehension.

Carrell (1987) not only studied the effect of topic familiarity and
text structure on reading comprehension but also examined their
relation to each other. Fifty-two ESL students were divided Into two
groups based on their religious background. Group 1 consisted of 28
students of Muslim background, and group 2 consisted of 24 students
of Roman Catholic background. Two reading passages were developed.
Each passage was a biography of little-known Muslim and Catholic
characters. Two versions of each passage were created; one version
was a well-formed historical narrative, while the other presented a
altered rhetorical organization. As predicted, the content
familiarity and familiar rhetorical structure condition yielded good
reading comprehension. On the other hand, unfamiliar content and
unfamiliar rhetorical structure yielded poor reading comprehension.
Carrell claims that when the subjects read a 'mixed' text (familiar
content, unfamiliar rhetorical structure; unfamiliar content,
familiar rhetorical structure) content familiarity affected
comprehension to a greater extent than rhetorical structure. it is
not explained, however, how the author defined topic familiarity.

Religion is such an important part of life in some social groups that

43

It is hard to understand in this study why religious background was
considered topic knowledge and not cultural background. _

Although results from research indicate that background
knowledge facilitates lnferencing and text comprehension, it is not
yet understood to what extent prior knowledge overrides language
difficulties. Research has not provided conclusive results.
Hammadou (1991) conducted two parallel studies, one in French, and
one in ltalian. Participants in the study were all native speakers of
American English enrolled in French and Italian courses. Students
were divided Into two categories: beginners and advanced. All
participants read authentic passages on familiar and unfamiliar
topics. Familiar and unfamiliar topics were selected based on a
previous survey of similar university students. The ranking for topic
familiarity of French passages was: 1) AIDS; 2) Stealth bomber; 3)
Sudan. The best recalled French passages were: 1) Stealth bomber; 2)
Sudan; 3) AIDS. The ranking by topic familiarity of Italian passages
was: 1) Sports; 2) Cinema; 3) Biography. However, the ltalian
passages in order of best recall were: 1) Biography; 2) Cinema; 3)
Sports. Results indicate that the topics judged as most familiar
(Sports and AIDS) were the least well recalled. Hammadou claims
that a possible explanation for her results may lie in the fact that
there is not a shortcut method for identifying prior knowledge. She
calls into question a subject's self report on her/his prior
knowledge. In her study, familiarity ranking was unable to predict
better recalls of the passage. An important issue that is not

considered in this study is the linguistic proficiency of the

44

participants. Although some of the students were considered
advanced compared to the other participants, the language used in
the unedited passages may have been beyond their linguistic
abilities.

Language proficiency and text features have also been studied as
major variables in the process of reading in a foreign language.
Clarke (1988) examined the reading behaviors of Spanish speakers
when reading in Spanish and in English. The participants were
Identified as good or poor L1 readers. Their answers on cloze tests
in Spanish and in English were analyzed. Results indicated that good
readers appeared to rely more on semantic cues than on syntactic
cues compared to poor readers when reading in their native language.
The gap between good and poor readers was reduced when the
subjects were reading in English; the use of syntactic cues was
equal by both good and poor readers when reading in the second
language. Clarke claims that although good readers may transfer
their L1 reading skills when reading in a foreign language, the
linguistic proficiency of the readers has an important role in text
comprehension. It seems that the reader's limited competence can
“short-circuit“ the good reader's system. This results in good
readers using poor readers‘ strategies because of their inability to
transfer successful L1 reading strategies.

Alderson and Urquhart's (1988) study supports the importance of
linguistic competence in reading comprehension. They found that the
background knowledge hypothesis was only partially confirmed, but

there was strong evidence that suggested that readers with higher

45

language level outperformed those readers with weaker language
knowledge. .

Few second language reading studies have investigated how
syntax at the sentence and text level affects reading comprehension.
This is an Important omission. Lack of data on this issue is rather
surprising since a considerable amount of time is spent In
second/foreign language classrooms explaining and practicing
syntactic structures (Bernhardt, 1991). Guarino and Perkins (1986),
Bhatia (1984), and Blau (1982) found that syntactic complexity does
not affect reading comprehension significantly.

More recently, results from Yano, Long, and Ross's (1994) study
indicate that linguistically simplified texts did not result in
significantly increased comprehension. The participants in this
study were 483 Japanese college students who read one of three
versions of 13 reading passages. Research indicates that native
speakers (NS) make various adjustments when conversing with
nonnative speakers (NNS) in order to facilitate comprehension. (For a
review, see Larsen-Freeman and Long, 1991; Long, 1983a.) Based on
these findings, Yano et al. (1994) rewrote 13 passages from a Wide
range of areas: biology, computer science, ecology, geography,
history, linguistics, and sociology. There were three versions of
each passage: (a) native baseline, (b) simplified, and (c) elaborated.
in the simplified form of the NS baseline version, sentences were
shorter and the number of embedded clauses was kept to a minimum.
in the elaborated form, paraphrasing was used to restate ideas and

to provide definitions of low-frequency content words. Results

46

indicate that comprehension was higher for those participants who
read the simplified version of the text, but not significantly higher
than it was for those who read the elaborated version. Yano et al.
claim that elaborated reading texts expose L2 reader to natural
discourse. Elaborated texts are semantically rich and help readers
make inferences and achieve better comprehension.

Strother and Ulijn (1987) provide consistent findings in this
regard. They found that there were no significant differences
between the subjects who read a syntactically simplified journal
article and those who read an unsimplified version of the same
article. Strother and Ulijn suggest that knowledge of vocabulary is
more important than syntactic complexity. They conclude that
“lexical rewriting“ rather than “syntactic rewriting“ of professional
texts will result in better comprehension.

The importance of vocabulary knowledge In the process of reading
has been well recognized in L1 reading. L1 reading research has
reported strong correlations between word knowledge and
comprehension (Wittrock, Marks, & Doctorow, 1975; Anderson &
Freebody, 1981; Nagy, William, Herman, & Anderson, 1985; Medo &
Ryder, 1993). Very few studies in L2 reading, however, have
investigated how vocabulary knowledge relates to comprehension.
This lack of data is rather surprising when we reflect that L2
reading research has been heavily dependent on L1 reading findings.
in addition, because reading In a second or foreign language involves
learning a new linguistic code, lower-level processing skills such as

word knowledge can be assumed to be critical. One possible

47

explanation for this lack of data may the strong influence that
schema theory and Goodman's psycholinguistic model have had on L2
reading research.

At the word recognition level, some L2 reading researchers have
examined how orthography affects reading. Koda (1987, 1992), and
Shimron and Sivan (1994) investigated the effect of the
orthographic system on the process of reading and word recognition.
Koda used Japanese students reading In English, and American
college students reading in Japanese as a foreign language. in both
studies, she found evidence supporting the claim that L2 readers
focus strongly on visual information present in graphemic features
when reading in a different orthographic system. This heavy
involvement in lower-level processing strains the reader's short-
term limited capacity, which frequently results in poor
comprehension. ln Shimron and Sivan's study, twenty four native
speakers of Hebrew and twelve native speakers of English read two
passages, one in Hebrew and the other In English. All the
participants were highly advanced bilinguals who used their second
language daily at work. Results of this study support the claim that
orthography is an important variable in the process of reading.
Shimron and Sivan's findings indicate that the text was read faster
In English than in Hebrew; even by native speakers of Hebrew.

Kim (1995) examined the effect of vocabulary and prior
knowledge on EFL students' recall protocols. The participants in this
study were 108 Korean students who had been studying English as a

foreign language for at least four and a half years. The students

48

were randomly assigned to one of the four treatments: (A) difficult
vocabulary without prereading instruction; (B) easy vocabulary
without prereading instruction; (0) difficult vocabulary with
prereading instruction; and (D) easy vocabulary with prereading
instruction. The prereading instruction was assumed to give the
students information about what the text content was about, and
knowledge in the content area of the text. Prereadlng Instruction did
not seem to have a strong effect on the amount of information
recalled. The means were not noticeable different between the
groups that read the difficult version, with or without prereading
instruction, and between the groups that read the easy version, with
or without prereading instruction. Results from this research
indicated, however, that prereading instruction influenced the type
of information that the participants included In their recalls.
Participants who read the difficult vocabulary text and received
prereading instruction reported more extratextual information,
“which mostly reflected blind reliance on what the subjects
remembered from the prereading instruction“ (p. 53). The results of
this study supported the claim that difficult vocabulary places a
certain strain on the students‘ comprehension capacity, and that
prereading instruction was not enough to help the students cope
with difficult vocabulary.

Most L2 reading studies that have focused on vocabulary have
been concerned with the pedagogical aspect of vocabulary knowledge
and reading. Having recognized the important role of vocabulary in

first language reading, L1 and L2 reading specialists have examined

49

techniques that teachers can effectively use to help their students
retain words. Some of the techniques that are often suggested are:
possible sentences (More and More, 1986), list-group-label
(sometimes also called semantic mapping ; Johnson and Pearson,
1984), contextual redefinition (Cunningham, Cunningham, and Arthur,
1981), feature analysis (Johnson and Pearson, 1984, and keyword
method (Pressley, Levin, and Delaney 1982). Further discussion of
these techniques is beyond the scope of this review.

The technique that has received the most attention in the L2
reading field is the keyword method. Results from some early
studies indicate that the keyword method facilitates the acquisition
of new words in a foreign language (Atkinson and Raugh, 1975;
Pressley, Levin and Delaney, 1982). There are two steps when
students use the keyword method, the acoustic stage and the image
stage. First, students find a known word that sounds like (and
possible looks like) the target word. Tierney, Readance, and Dishner
(1990) provide an example of this first stage when reading in
English as their first language. Students encounter the word potable
and think of the word pot that they already know. Both words are
relatively similar in sound and spelling. When reading in a foreign
language, students think of a word that may sound and look similar
in their first language. Native speakers of Spanish trying to
remember the meaning of the English word glob might think of the
Spanish word globo, which means balloon and is similar in spelling
and in sound. in stage two, the students create a visual image and

associate it with the target. For instance, after the word potable is

50

related to the word pot, the students think of a pot filled with
crystal clear, cool water (Tierney et. al, 1990). The Spanish
speakers may create the image of something round similar to a drop
that can be big or small. This technique Is believed to help them
store and retrieve words, so that next time native speakers of
Spanish read a phrase such as ‘a glob of paste,‘ or 'a glob of mashed
potatoes,‘ they will easily access the meaning of the English word.

More recently, Hogben and Lawson (1994) examined whether
experienced foreign language learners can also benefit from the use
of the keyword method. in a brief report of their study Hogben and
Lawson Indicate that this method can also be successfully used with
advanced foreign language learners.

One of the disadvantages that I can see when using this technique
is that it is time consuming for the teacher if snie generates the
keywords, or for the class if the teacher intends to have the
students come up with their own keywords. In the second stage, the
creating of the visual image, the teacher can provide the image that
links the key word and the target word; in this case there is no
personal involvement by the students, resulting in a potentially
extraneous image. On the other hand, if the students are asked to
generate their own image, it is likely that a wide variety of visual
Images will be created. This will make virtually impossible for the
teacher to help the students recall a word later on by referring to

their mental Image.

51

2.3 Vocabulary Knowledge and Reading Comprehension

Results from L1 reading research provided robust findings
indicating that vocabulary knowledge affects reading comprehension.
In fact, vocabulary knowledge have been found to correlate with text
comprehension (Anderson and Freebody, 1981). Since research also
indicated that prior knowledge facilitates comprehension, several
L1 reading studies were carried out to investigate the compensatory
hypothesis presented by interactive models of reading (Rumelhart,
1977; Stanovich, 1980, 1984). According to this hypothesis,
deficiencies at any level during the reading process can be
compensated for at another level. in other words, a reader that has
difficulties at lower-level process, can compensate for this
deficiency by using high-level processing. Results from these
studies failed to provide evidence of such interaction between
vocabulary knowledge and prior knowledge (Freebody and Anderson,
1983b; Kim and Goetz, 1994; Medo and Ryder, 1993; Stahl, Hare,
Sinatra, and Gregory; Stahl and Jacobson, 1986; Stahl, Jacobson,
Davis and Davis, 1989).

Stahl et al. (1989) studied how vocabulary difficulty and prior
knowledge affect reading comprehension of an unfamiliar text. Their
findings are consistent with results from earlier studies. ‘
Vocabulary and prior knowledge seem to affect recall, but these
factors affect comprehension in different ways. Those students who
read the difficult vocabulary version recalled more information out
of order. it appears that unknown vocabulary hampers the

development of a coherent text. The authors concluded that

52

vocabulary and background knowledge function independently and
that readers do not use one higher-level process, such as prior
knowledge, to compensate for problems in another higher-level
knowledge source, such as difficult vocabulary.

Stahl (1991) suggests that the ongoing comprehension process
has two components, microprocessing and macroprocessing.
Microprocessing involves the development of a microstructure; this
implies the combination of propositions into a coherent text. This
process allows readers to recall coherent and ordered ideas from the
text. Macroprocessing involves the development of the gist of the
text. Readers can get a general idea of the text but have difficulty in
understanding the relationship among the ideas presented.

As in first-language studies in reading, the importance of
vocabulary has also been recognized In L2 reading research (Koda,
1989; Krashen, 1989; Laufer, 1990; Schulz, 1983). However, whereas
current L1 reading researchers continue to examine the role of
vocabulary in the process of reading, this interest in vocabulary is
not so strong in L2 reading studies. Stoller and Grabe (1993) claim
that “current L1 research suggests that vocabulary is much more
central to understanding processes than L2 theorists and
practitioners have generally recognized.“ (p. 25)

Eskey (1988) recognized the importance of background knowledge
in reading comprehension, but also warned us of the potential
hazards of de-emphasizing “the perceptual and decoding dimensions

of the process of reading. (p. 93)

53

Conclusions

Although most foreign language teachers have always perceived
vocabulary knowledge as a predictor of success or failure in L2
reading, they have not received much evidence for this from
empirical research. The strong Influence of the psycholinguistic
model and schema theory for over a decade have resulted in a large
number of studies that focused on top-down processing.

Despite robust evidence provided by L1 reading research on the
important role that vocabulary knowledge plays in the process of
reading, very few studies have examined how vocabulary knowledge
affects the process of reading in a foreign language.

There is some evidence that indicates that poor lower-level
skills may 'short-circuit' the reading process in a foreign language,
which will. result in poor text comprehension. The orthographic
system has proven to affect the process of reading. Although there
is some research that focused on some bottom-up recognition skills,
no studies have directly investigated how unknown words affect
comprehension, and what aspects of the process are affected the
most by unknown vocabulary. it is easy to understand the need of
empirical research on the effect of vocabulary on the L2 reading
process. Results from such studies will contribute to the
development of second language reading models. Sound evidence from
empirical research will also help reading teachers develop effective
techniques to help language learners become efficient and effective

readers.

CHAPTER THREE

WY

The purpose of this study is to investigate the effects of
background knowledge and vocabulary on the process of reading in a
foreign language, and the possible interaction between these two
factors. Several studies carried out during the 19803 demonstrated
the importance of background knowledge in the process of reading in
a second language (Alderson and Urquhart, 1988; Carrell, 1983a, b,
1984b; Carrell and Eistehold, 1983; Johnson, 1982, Lee 1986). Since
most foreign-language readers do not have an extensive knowledge
of the.forelgn culture, peoples' customs and beliefs, the role that
background information has in reading in foreign-language may be
even more important than it is in first language reading.

The importance of vocabulary knowledge In the process of reading
has been well recognized in first language (L1) reading. Studies in L1
reading have reported that there are strong correlations between
word knowledge and reading comprehension (Wittrock, Marks and
Doctorow, 1975; Anderson and Freebody, 1981; Nagy, William, ,
Herman and Anderson, 1985; Medow and Ryder, 1993). Even though
studies on the process of reading in L2 have closely followed L1
reading research in general, the interest In vocabulary Is not
parallel between the two fields. Whereas L1 reading researchers
continue to investigate the role of vocabulary In reading
comprehension, such research in L2 reading has been scarce. There

54

55
are a few early studies that emphasized the importance of

vocabulary in ESL reading (T waddell, 1973; Eskey, 1973) but most
recent studies have focused on examining how prior knowledge
affects reading. in some cases researchers examined the
relationship between overall language proficiency and reading
comprehension (Clarke, 1988; Alderson and Urquhart, 1988). But in
such studies researchers concentrated more on the effect of syntax
rather than the effect of vocabulary when analyzing their data.
During the 19805, data from L1 reading research indicated that
these two factors-background knowledge and prior knowledge--are
closely related to each other. in fact, Stanovich (1980, 1984)
claimed that there is an Interactive—compensatory effect between
prior knowledge and vocabulary during the process of reading.
According to this hypothesis, L1 readers who have difficulties at
one level may compensate for this deficiency by relying more on
another level. in any particular reading situation, for instance,
readers who find several unknown words can still understand the
text if the topic is familiar to them. Research In L1 reading,
however, has failed to provide strong evidence to support the
interactive-compensatory hypothesis. On the contrary, results from
some L1 reading studies indicate that the two factors-background
knowledge and vocabulary--function Independently. They may both
affect comprehension but in different ways. Background knowledge
seems to affect macroprocessing while vocabulary appears to affect
microprocessing. Macroprocessing allows the reader to get the gist

of the text. Microprocessing allows the reader to combine the words

56
into propositions and involves the combination of those propositions

into coherent texts.

This study will Investigate how prior knowledge and unknoWn
vocabulary affect students' written recall of central and supporting
ideas and the effect of these two variables on comprehension in a
multiple-choice task. The possible interactive effects of background
knowledge and vocabulary vw‘li also be examined. This research will
attempt to answer the following questions:

a) How does prior knowledge affect information recall?

b) How does vocabulary difficulty affect information recall?

c) How does prior knowledge affect comprehension?

d) How does vocabulary difficulty affect comprehension?

9) Do prior knowledge and difficult vocabulary function

interactively?

The following hypothesis regarding the role of vocabulary and
background knowledge in L2 reading were formulated:

1. L2 readers who read about a familiar topic will perform
better on the recall of central units than those who read
about an unfamiliar topic.

2. L2 readers who read about a familiar topic will perform
better on the recall of supporting units than those who read
about an unfamiliar topic.

3. L2 readers who read an easy vocabulary text will perform
better on the recall of central units than those Who read a
difficult vocabulary version.

4. L2 readers who read an easy vocabulary text will perform
better on the recall of supporting units than those who read a
difficult vocabulary version.

10.

11.

12.

13.

14.

15.

57

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity in their recall of central units

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity in their recall of supporting units.

L2 readers who read about a familiar topic will not perform
better on answering textually explicit questions than those
who read about an unfamiliar topic.

L2 readers who read an easy vocabulary text will perform
better on answering textually explicit questions than those
who read a difficult vocabulary version.

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity to answer textually explicit questions.

L2 readers who read about a familiar topic will not perform
better on answering textually implicit questions than those
who read about an unfamiliar topic.

L2 readers who read an easy vocabulary text will not perform
better on answering textually implicit questions than those
who read a difficult vocabulary version.

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity to answer textually implicit questions.

L2 readers who read about a familiar topic will perform better
on answering scriptically implicit questions than those who
read about an unfamiliar topic.

L2 readers who read an easy vocabulary text will not perform
better on answering scriptically implicit questions than those
who read a difficult vocabulary version.

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity to answer scriptically implicit

58
quesfions

3.1 SUBJECTS

Participants In the study were 105 high-school students in five
intact classes in a mid-size town in the midwestern United States.
All the participants were juniors and seniors who consented to
participate in the study. Their participation in the study was
entirely voluntary and no student received any monetary
compensation. it was clearly explained to them that the activities
they were going to do during the study would not bring any changes
into the classroom or in the way their classes were usually
conducted. It was also explained to them that their performance on
the activities was not going to affect their grades In any way. All
the explanations were given in English to be sure that all the
students understood them.

From the original 105 participants, the responses of 62 students
were considered in this study. Responses from 43 students were not
Included because either (a) Spanish was the language used at home,
or (b) they had been studying Spanish as a foreign language for less
than four years, or (c) they did not attend all the classes, when the
study was explained and the data collected. Table 1 presents the
number of students in each class whose responses were not included

In the study and the reasons for not including them.

59

TABLE 1
STUDENTS“ RESPONSES NOT INCLUDED IN THE STUDY

 

Group Number of students Reasons
whose responses
were not included

 

Spanish used at home

 

missed some readings

 

missed some readigs

 

Studied Spanish less than 4 yrs.

 

missed some readings

 

studied Spanish less than 4 yrs.

 

missed some readifls

 

missed some readiqus

 

ranOOCDtD>>
\lOJNVb-VOJOJN

 

 

 

studied Spanish less than 4 yrs.

 

 

Except for one student who was 14 years old, all the participants
in the study ranged in age from 15 to 18. The fourteen-year-oid
student's responses were included In the study because they were

not significantly different from those provided by older students.

TABLE 2
AGE AND GENDER OF PARTICIPANTS

 

 

 

 

 

 

 

 

 

__A_ge 18 17 16 15 14
Females 6 19 9 4 1
Mates 3 9 10 1 0

 

Of the 62 students whose responses were included in the study,

thirty nine were females and twenty three males. (Table 2)

 

60
Sixty students indicated that English was the language spoken at

home. One student reported that he used Persian at home and another
student reported that Bangladeshi was the language spoken at home.
According to their classroom teacher neither of these students

performed differently from the rest of the class in regular

 

 

 

 

 

 

activities.
40
30-4
to
E
o
i
.5 20-
i
z 101
Yearsof
study
0

 

 

 

 

FIGURE 3
NUMBER OF YEARS THAT THE STUDENTS HAD STUDIED SPANISH

.All the students whose responses were included in this study had
been studying Spanish as a foreign language for a period of 4 to6
years. Thirty-two students had been studying Spanish for 4 years;
twenty students had studied the language for 5 years and ten
students for a period of 6 years. Figure 3 shows the participants
grouped according to the length of time they had been studying
Spanish.

61
Twenty-four participants had visited a Spanish speaking country.

Ten students had visited Mexico; seven had been to Spain; 6 had
visited Costa Rica and one student had been to El Salvador. The
length of time they stayed in the Spanish speaking countries varied
from 2 days to 6 weeks. Figure 4 represents both groups of students,
those who had never visited a Spanish speaking country and those
who visited countries where Spanish was the native language. Figure
5 represents the Spanish speaking countries that the students

visited.

Yes
No.

 

FIGURE 4
STUDENTS WHO VISITED SPANISH-SPEAKING COUNTRIES

Mexico

fig!

Spain

Costa Rica

El Salvador

 

FIGURE 5
SPANISH-SPEAKING COUNTRIES VISITED BY THE STUDENTS

Figure 6 represents the length of time those students stayed in
the different Spanish-speaking countries. Seven students stayed in
the Spanish-speaking country they viated for a week or less; four
students stayed in the country for two weeks; six students spent 3
weeks, and seven students stayed In the Spanish-speaking country

for four weeks or more.

63

 

 

four weeks or more

 

 

 

izizizizizizizizi:3:1:1:3:3:3:3:1:1:3:3:1:3:1:3:3:3:3:1:1:3:1:1:235:23: 9’
three weeks 35:3:3:3:3:1:3:1:1:3:1:3:3:1:1:123:3:3:3:3:1:1:3:=:3:3:1:3:3:1:1:1:3:1:1:3:

 

two weeks

 

Number of
students

a week or less

 

 

 

 

0
21
4
6
8..i

FIGURE 6
LENGTH OF STUDENTS' STAY IN A SPANISH-SPEAKING COUNTRY

Some of the students who participated In the study had Spanish
speaking relatives or friends. Thirty-one participants indicated that
they did not have any relatives or friends who spoke Spanish.
Thirteen students Indicated that their Spanish speaking friends used
Spanish only when they talked to each other, but not when they
talked to the participants of the study. Eighteen students said that
their Spanish speaking friends used Spanish not only when they
talked to each other but also when they talked to the students.
Figure 7 represents the students who did not have Spanish speaking
relatives or friends and those who had Spanish speaking friends. The
section that represents the students who had Spanish speaking
friends/relatives also indicates those participants whose

friends/relatives spoke Spanish to them and those whose

64
friends/relatives used the language while the students were around

but not when talking to the participants In the study.

Students who did not have
Spanish-speaking friends
or relatives

Students with Spanish-speaking
friends who used Spanish when
talking to each other

I Students whose Spanish-
speaking friends used
Spanish when talking to
them

 

FIGURE 7
STUDENTS WHO HAD/DID NOT HAVE SPANISH-SPEAKING FRIENDS

Forty-five students reported that they were exposed to Spanish
outside the classroom by watching television, and reading books and
magazines in Spanish. The rest of the students said that their only
contact with the foreign language was in the classroom.

The majority of the students reported that they were planning to
continue studying Spanish after they finish high school. Forty-five
participants indicated their interest in continuing with the study of
the language beyond high-school; seven students said that they were
not sure at that time, and ten participants indicated that they had
decided not to continue teaming Spanish after they graduated. Figure

8 represents the students who were in contact with Spanish outside

65
the classroom and those who were not. The same figure indicates the

number of students In each group (exposed and not exposed outside
the classroom) that planned/did not plan to continue studying '
Spanish after they finished high school.

 

.9

C

.9

3s

8%

00.

gm

22'

n.—

os lYes
g3

mm Notsure
‘sé

3's I No
(‘58

 

 

 

 

Students exposed/not exposed to Spanish outside
the classroom

FIGURE 8
STUDENTS' LANGUAGE EXPOSURE AND PLANS TO CONTINUE STUDYING

3.2 MATERIALS

The nature of the study required that special attention be paid to
the topics selected for the reading texts. Two passages were used:
one reading passage was on a familiar theme, the other presented an

unfamiliar topic. Since all the participants in the study were

66
assumed to be American high-school students2 , Halloween was

selected as the topic for the familiar passage. The reading passage
on an unfamiliar topic was about La Diablada, a festival'held in
Bolivia every year. La Diablada was considered an appropriate topic
because: a) Bolivia is a Latin American country with which not many
high-school students are familiar; b) La Diablada is a festival that
has not been highly publicized like other South American festivals,
such as Carnival in Brazil.

Two original reading passages were developed for the study. The
author examined different reading materials and tests that the
students had used in class to get an idea of what would be
linguistically suitable for the participants. Then, two passages were
constructed, one about Halloween and the other about La Diablada.
After the reading passages were written, they were read by two
native Spanish speakers to make sure that the reading sounded
natural. One of the Spanish speakers had formal linguistic training
and experience in teaching Spanish as a foreign language in the
United States). The other Spanish speaker had no linguistic training
and had no experience in teaching any languages. His area of study is
Biochemistry, and he had been living in the States for just a year at
the time he read the passages. The changes suggested by both _
Spanish speakers were discussed and incorporated. There was total
agreement between the Spanish speakers themselves and the author
on what became the final version of the reading passages. The

passage about Halloween consisted of three paragraphs; the easy

 

2 Two students who participated in the study were not Americans; one came from Iran
and the other from Bangladesh. Both of them, however, had lived in the United States for
over five years, long enough to be familiar with Halloween.

67
version contained 269 words and the hard version had 267 words.

The passage about La Diablada consisted of three paragraphs; the
hard version contained 290 words and the easy version contained
289 words.

Based on the two original passages, two versions of each
passage were developed, a difficult version and an easy version. The
level of difficulty was determined by the vocabulary used in the
texts. The easy version contained words frequently used in
intermediate/high Intermediate reading materials. The difficult
version contained words that are generally used less frequently. The
reading passage that contained these words was considered the hard
version of the reading passage. The procedure used to create the
easy and the difficult versions of each reading passage was modeled
on the procedure used by Stahl, Jacobson, Davis and Davis (1989) in
their study on the effect of prior knowledge and difficult vocabulary
when reading in the first language. The easy version included words
that were thought to be known to the students. As will be discussed
later in this chapter, there were three words in both versions (easy
and hard) of the passage about La Diablada that were considered
easy. As a result of the post-data collection activity, it was found
that three participants did not know these words. For the hard
version, approximately every sixth content word was replaced by a
difficult (unknown) synonym. Special attention was paid to make
both versions sound natural, avoiding the use of stilted language. in
some cases, a function word was changed, added or deleted in an
effort to make the modified text sound natural. After the two

versions of each reading passage were developed, they were read by

68
the two native Spanish speakers who read the original passages to

make sure that the four final versions did not sound affected or
artificial. Excerpts from the two versions of each passage are'shown
in Table 3.

TABLE 3
SAMPLE PARAGRAPHS FROM EASY AND HARD TEXT VERSIONS

 

_ —— ——————.—-—-—-——u_————————————o—_———u——~————u———————

Familiar topic - easy version

Una de las fiestas mas esperadas por los nifios de Norteamérica es la
famosa ”Noche de Brujas". Ese dia por la tardecita, los nii‘ios van de
puerta en puerta visitando amigos y vecinos. Los nifios y a veces
tambien los adultos llevan mascaras, algunas muy teas, que
representan personajes de ciencia ficcién, figuras del gobiemo y
muy especialmente brujas, fantasmas y esqueletos.

Familiar topic - hard version

Una de las fiestas mas esperadas por los nifios de Norteamérica
es la famosa “Noche de Brujas". Ese dia por la tardecita, ios nihos
van de puerta en puerta visitando amigos y vecinos. Los nifios y a
veces tambien los adultos llevan mascaras, algunas horripiiantes,
que representan personajes de ciencia ficcibn, figuras
gubemamentales, y muy especialmente brujas, fantasmas y
esqueletos.

 

 

 

Unfamiliar tepic - easy version

Cada afio se celebra en Oruro, Bolivia, la famosa “Diablada'. Esta
es una de las fiestas mas originates e interesantes de Sudamértca.
La poblacién de Oruro esta compuesta, en su mayor parte, por indios.
Todos los anos en tiempo de Camaval, en febrero 0 marzo, estos
indios abandonan sus peligrosos trabajos en las minas y se ponen a
bailar al ritmo de las guitarras y otros Instrumentos tlpicos. Los
bailarines llevan mascaras que representan diferentes figuras de la

69

TABLE 3 (cont'd)

historia y de la mltologia, asl como también sapos, serpientes'y
especialmente diablos, centenares de diablos. Muchos de los turistas
que llegan a ver la diablada no entienden ei doble significado de la
fiesta.

Unfamiliar topic - hard version

Cada a‘r'to se celebra en Oruro, Bolivia, la famosa “Diablada". Esta
es una de las fiestas mas interesantes de Sudamérica. La poblacién
de Oruro esta compuesta, en su mayor parte, por quechuas. Todos los
ahos en tiempo de Camaval, en febrero 0 marzo, estos indios dejan
sus peligrosos trabajos en las minas y se ponen a bailar al ritmo de
los charangos y otros instrumentos tlpicos. Los bailarines llevan
caretas que representan figuras de la historia y de la mltologia, asl
como también batracios, serpientes y especialmente diablos,
centenares de diablos. Muchos de los turistas que llegan a presenciar
Ia diablada no entienden el doble significado de la fiesta.

 

 

The first two sentences and the last sentence were the same in
both versions-easy and hard--of the two reading passages on the
same topic. A copy of both reading passages-familiar and
unfamiliar--hard and easy versions, along with the translations into
English are included in Appendix A.

3.3 lNSTRUMENTS

The written-recall measure. The first measure used after the
participants read the passages was a free written recall. After
reading the passage, each student received a blank page with

instructions printed at the top. The instructions, written in English,

70
asked the participants to write as much Information as they recalled

from the passage they had just read.

in order to quantitatively assess the readers' protocols, each
reading passage had been divided into linguistic units. in addition to
the two native speakers of Spanish who had read the first versions
of the reading passages, two other near native speakers of English
worked independently to divide the passages into clauses following
Lee's (1987) procedures adapted from Johnson (1970). Johnson was
the first to devise an objective method for dividing prose. Lee
(1987) adopted Johnson's methodology and suggested dividing the
text into clauses. Table 4 shows how the first paragraphs of the
easy version of the reading texts were divided.

All disagreements were discussed and resolved before the

data were collected. The passage about Halloween was divided into
twenty-four linguistic units, and the passage about La Diablada was

divided into twenty-seven.

TABLE 4
SAMPLE PARAGRAPHS FROM FAMILIAR AND UNFAMILIAR TEXTS

La Noche de Brujas

 

IUna de las fiestas mas esperadas por Ios ninos de Norteamerica
es la famosa "Noche de Brujas“. I Ese dia por la tardecita, los nii‘tos
van de puerta en puerta visitando amigos y vecinos. I Los nii'ios y a
veces tambien los adultos llevan mascaras, I algunas muy teas, I
que representan personajes de ciencia ficcién, figuras del gobiemo y
muy especialmente brujas, fantasmas y esqueletos. I

La Diablada
ICada afio se celebra en Oruro, Bolivia, la famosa "Diablada“. I

71

TABLE 4 (cont)

Esta es una de las fiestas mas originales e interesantes de
Sudamérica. I La poblacién de Oruro esta compuesta, en su mayor
parte, por indios. I Todos los anos en tiempo de Camaval, en febrero
0 marzo, estos indios abandonan sus peligrosos trabajos en las
minas / y se ponen a bailar al ritmo de las guitarras y otros
instrumentos tipicos. I Los bailarines llevan mascaras I que
representan diferentes figuras de la historia y de la mltologia, I asl
como también sapos, serpientes y especialmente diablos, centenares
de diablos. I Muchos de los turistas que llegan a ver la diablada I no
entienden el doble significado de la fiesta.

 

Modeled on the procedure suggested by Johnson (1970), the
clauses were then classified according to their importance to the
whole text. When the importance of the clauses is taken into
consideration, they can be classified as central, supportive, and
distracting. Using this classification, the same readers who divided
the text into clauses were then asked to rank the clauses according
to their importance to the passage. This was done in two steps. First
the readers were asked to cross out clauses until only half of the
original number of clauses was left. They were instructed to cross
out those clauses which 'could be eliminated without destroying the
essence of the story" (Johnson, 1970 p.13). At a second stage, the
readers were presented with the clauses they had left and asked to
cross out half of them so that only those clauses that constituted
the core of the reading remained. The units left were classified as
central clauses and the units that were crossed out during this

second stage were considered supportive clauses. All the clauses

72
that were crossed out the first time were classified as distracting

clauses.

As a result of this classification, six linguistic units from 'the
passage about Halloween were classified as central units and six as
supporting units. From the passage about La Diablada, seven
linguistic units were classified as central units, and six as
supporting units. Table 5 and 6 present the central, and supporting

units from the reading passages.

TABLE 5
CENTRAL AND SUPPORTING UNITS FROM HALLOWEEN

 

Central units:

-La Noche de Brujas se inicio como un festival Celta
-(Los Celtas) dedicaba el Ultimo dia de octubre al Senor
de la Muerte

-(Para los Celtas) en noviembre empezaba el invierno, la
estacién del frio, la tristeza y la oscuridad

-Los Ceitas asociaban el primer dia de noviembre con los
muertos

~(Fue) durante la dominacién romana

-(que) la celebracién celta se mezclo con la cn‘stiana

Supporting units:

-(La Noche de Brujas) es una de las fiestas mas
esperadas

-Los nii’ios y los adultos llevan mascaras

-(Las mascaras) representan figuras del gobiemo y
fantasmas, brujas y esqueletos

-La gente encendia fuego por la noche

-para asustar a las malas influencias

-El Papa declare el primero de noviembre el Dia de Todos
los Santos

—_———_—-—-—_-——-—————-————c——————o—_———-—.—————.-_—_——————_———-———

73

TABLE 6
CENTRAL AND SUPPORTING UNITS FROM LA DIABLADA

Central units:

-Cada afio se celebra en Oruro, Bolivia, la famosa
Diablada

-Cuenta la leyenda de los Andes que hace mucho tiempo,
el antiguo pueblo andino de los Urus lievo una vida de
mucha libertad

-(luego) abandono las malas costumbres

-Un espiritu malo resolvio vengarse

-Entonces aparecio Nusca, una princesa lnca, que
transformo

a ios animales en piedra.

-Para muchos habitantes de la zona y para la mayor parte
de los indios, la Nusca es la Virgen de Socabén.

-(La Slusca) es la figura central de La Diablada

Supporting units:

-Todos los anos en tiempo de Camaval, en febrero 0
marzo, ios indios abandonan sus peligrosso trabajos en
las minas

-Los bailarines (ios indios) llevan mascaras

-que presentan diferentes figuras de la historia y la
mitologia

-Un dia (el pueblo de los Urus) cambio su manera de vivir
-(Un espiritu malo) trato de vencer a los Urus
-enviandoles muchos y diferentes animales

 

 

The comprehension measure. The measure used to assess
students' comprehension of the texts was a multiple-choice test.
Twelve multiple choice questions were developed for the familiar

text and twelve for the unfamiliar text. All the questions and the

74
options were In English. The procedure used to write the questions

was modeled on the procedure used by Stahl, Jacobson, Davis and
Davis (1989) In their study on the process of reading in the first
language. According to their coding scheme, four of the questions
were textually explicit; four were textually implicit; and four were
scriptically implicit.

A question whose answer was explicitly stated in the text was
defined as a textually explicit question. A textually implicit
question was defined as one whose answer had to be inferred from
the text. A scriptically implicit question was defined as one whose
answer had to be inferred by the reader on the basis of his! her prior
knowledge; this definition was also used in Pearson and Johnson
(1978), Stahl and Jacobson (1986), and Stahl, Jacobson, Davis, and
Davis (1989). All the questions and possible answers were presented

in English in order to avoid comprehension problems.

3.4 PROCEDURES

The students who participated in this study had a fifty-minute
class in Spanish every day. The data for this study were collected
during three class periods. Although the classroom Instructor stayed
in the classroom during the time the study was carried out, all the
data were collected by the researcher.

On the first day, all the students were asked to fill out a
questionnaire that provided inforrnatlon about their age, gender,
language used at home, and length of time they had been studying
Spanish. in order to get more information about their exposure to the

language outside the classroom, they had to Indicate whether they

75
had lived in and/or visited a Spanish-speaking country; those IM'IO

answered any of these two question affirmatively were asked to
Indicate the length of their stay in that country. The students "were
also asked to indicate whether they had Spanish-speaking friends or
relatives who used Spanish when talking to them. The last two
questions were about whether they read or watched television In
Spanish and whether or not they were planning to continue studying
the language after they finished high school. After the students
completed the questionnaire and handed it in to the researcher, they
were asked to complete a consent form which asked for permission
to use the data collected for a study on the process of reading in a
foreign language. Copies of the questionnaire and the consent form
are included in Appendix A.

Thestudents were told that during the following two days they
were going to read two texts. They were also told that they could
read the texts at their own pace and as many times as they needed.
The participants were informed that there was no time limit for
reading each passage. They were encouraged to make an effort to
understand as much as they could because they were going to
perform some activities after each reading. in order to ensure that
all the students understood what they were going to do, all the
explanations and directions were given in English.

On the following two days, each group was randomly assigned to
read one of the two topics-familiar or unfamiliar. On day one of the
data collection, groups B, D and E read the text that presented a
familiar topic, Halloween; groups A, and C read the unfamiliar topic,

La Diablada. On the second day that the data were collected, groups

76
B, D, and E read the unfamiliar topic and groups A and 0 read the

familiar topic.

This resulted in all groups reading both reading texts-familiar
and unfamiliar. Participants stayed in their classes and were
randomly assigned to one of the two versions of the reading text--
hard or easy. A total of thirty-two participants read the hard
version of the unfamiliar topic, and thirty students read the easy
version. Thirty-one students read the hard version and the same
number of students read the hard version of the familiar topic. Table
7 represents the reading conditions for each group; figures in
parenthesis indicate the number of students that read the easy and

hard versions in each group.

TABLE 7
SCHEDULE OF READING CONDITIONS

 

 

 

 

 

DayOne Day Two
Group A La Diablada Halloween

(hard: 10; easy=7) (hard: 9; easy: 8)
Group B Halloween La Diablada

(hard: 7; easy: 6) (hard: 6; easE 7)
Group C La Diablada Halloween

(hard: 4; easy: 4) (hard: 4; easy: 4)
Group D Halloween La Diablada

(hard: 7; easy: 6) (hard: 6; easy: 7)

 

Halloween La Diablada
(hard: 4; easy: 7) (hard: 6; easy: 5)

Group E

 

 

 

 

 

77
The same data collection procedure used on day one was used on

day two. Students received their reading texts and took as much

time as they needed to read the passages. They were Instructed not
to use dictionaries, books, or consult with their classmates while
reading. When they finished, they returned the reading passage to the
researcher and received a blank sheet. The instructions on the sheet
asked them to write in English as much information as they recalled
from the text they had just read. A copy of the sheets where the
students wrote their free recalls for each reading text--familiar
and unfamiliar-is included in Appendix A.

When the students finished writing, they gave their written
recalls to the researcher. Following the free recall, participants
received twelve multiple-choice questions to assess their
comprehension. Each question had four options. No answer sheets
were used; participants marked their answers in the booklet. Four of
the questions were textually explicit; four were textually implicit;
and four were scriptically implicit. A copy of the multiple-choice
task for each reading passage is included in Appendix A.

Upon completion of the multiple-choice task, participants turned
It in to the researcher. Those students who finished first were
allowed to study or prepare homework for other classes. On the first
day of data collection, the classroom instructor let a few students
go to the school library.

On the second day, those students who finished first were also
allowed to do work for other classes but were asked not to leave the
classroom. it is important to stress here that this study was

designed to investigate the effect of prior knowledge and unknown

78
vocabulary on the process of reading. in order to evaluate the

participants prior knowledge about La Diablada and Halloween, and
verify assumptions about topic familiarity and easy and- hard
vocabulary, two activities were done on the second day of the data
collection. After all participants had finished the multiple-choice
task and returned it to the researcher, they were asked to rate
orally how familiar they were with the topics they had read. Three
options were written on the board and participants were asked to
raise their hands when they agreed with the option that was read.
The options on the board were presented in different order to
different groups. Table 8 presents how the options were presented
to the five groups. All participants in all groups Indicated that they
were 'very familiar' with the topic Halloween.

The ,same procedure was used to evaluate students' prior
knowledge about La Diablada. Results indicated that none of the
participants was familiar with La Diablada. All the students claimed

that they had never heard of or read about La Diablada.

TABLE 8
OPTIONS PRESENTED TO STUDENTS TO RATE TOPIC FAMILIARITY

 

 

 

QIQLILLA
The to ic
p HALLOWEEN is -very familiar to me
-a little familiar to me
-not familiar to me
65142.3
The topic HALLOWEEN is -not familiar to me
-a little familiar to me
-very familiar to me

 

 

 

79

TABLE 8 (con't)

 

 

 

W
The topic HALLOWEEN is -very familiar to me
-not familiar to me
-a little familiar to me
Group D
The topic HALLOWEEN is -not familiar to me
-very familiar to me
-a little familiar to me

 

 

 

Following this activity, each student received a copy of one of the
reading texts. Copies of both reading texts (familiar and unfamiliar)
and copies of both versions (hard and easy) were randomly
distributed among the students in every group. They were asked to
reread 'the passage and underline all the words they did not know. As
was predicted, all the hard words In both passages-Halloween and
La Diablada--were unknown to the students. in addition to the
predicted hard words, three more words were underlined by three
students in the passage about La Diablada (indios,' 'lugarehos,‘ and
'ritmo'). it is interesting to notice that two of these words are very
similar in English, 'indios‘ (indians), and 'ritmo' (rhythm--more
obvious in pronunciation than in spelling). Even though these words
seem to be transparent In meaning, three participants failed to
detect the connection between the Spanish and the English versions.
'Lugarefios' refers to the people who live in a particular area 'lugar.‘
Although 'lugar' could be considered a high frequency word, three
participants were not able to perceive the link between 'lugar' and

'lugarefio.‘ The fact that all these words were Included in the

80
unfamiliar text may have Influenced the way participants viewed the

words.

CHAPTER FOUR

ANALYSIS AND RESULTS

The first part of this chapter presents the quantitative analysis
of the data collected for this study. First, the effect of topic
familiarity and vocabulary on recall measure and comprehension
measure Is analyzed using descriptive statistics. Then a
multivariate analysis of covariance (MANCOVA), a multivariate
analysis of variance (MANOVA), and several univariate analyses of
variance (ANOVA) were applied to the participants' recall of
information, and to their comprehension scores.

ln,order to analyze the recall measure further, the participants'
written protocols were individually examined. The second part of

this chapter contains the results of the qualitative analysis .

4.1 QUANTITATIVE ANALYSIS
Statistical Analysis

Descriptive statistics. First, descriptive statistics were used to
analyze the two dependent variables-recall and comprehension. As
was expected, the mean scores were higher for recall when the
participants read a familiar text (familiar: X: 4.339, SD: 2.172;
unfamiliar: 5i: 3.323, SD: 2.141). Vocabulary difficulty also affected
recall scores. The mean for participants who read the easy version
(known vocabulary) was higher than for the students who read the
version that contained unknown vocabulary (easy: X: 4.393,

81

82
SD: 2.216; hard: 2: 3.286, SD: 2.075). A summary of descriptive

statistics for recall scores Is presented in Table 9.

TABLE 9
DESCRIPTIVE STATISTICS FOR RECALL SCORES

 

 

 

 

 

 

 

Variable bckmnd N Mean Median StDev SEMean
Recall 0 62 3.323 3.000 2.141 0.272
1 62 4.339 4.000 2.172 0.276
vcbdiff
0 63 3.286 3.000 2.075 0.261
1 61 4.393 4.000 2.216 0.284

 

 

Code: Bckgmd: prior knowledge; 0: unfamiliar topic; 1: familiar
topic.
Vcbdiff: vocabulary difficulty; 0: hard version/unknown
vocabulary; 1: easy version/known vocabulary.

 

 

 

Histograms are graphic representations used to show the
frequency distributions of the data. Histograms were also used in
this study to display graphically the frequency distribution of recall
scores. Figures 9 and 10 show the frequency distribution of recall
scores for familiar and unfamiliar texts. Figures 11 and 12 display
the frequency distribution of recall scores for hard and easy

versions.

83

 

N: 62 bclgmd: o

 

 

Midpoint Count
0 3 was
1 7 warrants
2 14 xxxxwftxwxxsswx
3 15 wwwsaaawxnwxxww
4 9 xxxxxwxxx
5 7 xxxxxxx
6 1 *
7 3
8 1 "
9 0
10 2 **
HGUREQ

RECALL SCORES FOR UNFAMILIAR TEXT

 

N: 62 mm: 1

 

 

Midpoint Count
0 1 *
1 5 warns
2 8 nsxxwxww
3 9 unsanitary:
4 13 wsxnnxxxsnxxrt
5 5 «was:
6 1o wwswwxrtxww
7 4 Kirk!
8 7 assassin
9 0
10 0
FlGUREfO

RECALL SCORES FOR FAMILIAR TEXT

84

 

 

 

 

 

 

 

 

N=63 vcbdiff: 0
Midpoint Count
0 4 wilt!
1 9 tuttttt‘kl‘
2 12 waxwxwwwwwsw
3 12 wants-warrants
4 10 thsanrtnx
5 7 «warns:
6 3 tin
7 3 xxx
8 3 kit!
9 0
10 0
FIGURE11
RECALL SCORES FOR HARD VERSION
N=62 vcbdiff:1
Midpoint Count
0 0
1 3 xxx
2 10 wwxxaawwww
3 12 wwnawwawwwrtrt
4 12 xwxxswxwwwwn
5 5 watts:
6 8 wasnxwxw
7 4 7(th
8 5 strains
9 0
10 2 *"
FIGURE12

RECAIJ. SCORES FOR EASY VERSION

85
The same procedure was used for students' comprehension scores,

and the results obtained were similar. The mean for comprehension
scores was higher for those participants who read the familiar topic
(familiar: X: 8.742, SD: 1.890; unfamiliar: X: 7.694, SD: 2.053).
Students' comprehension scores also showed differences. The mean
for comprehension scores was higher for participants who read the
passages that contained known vocabulary (easy: it 8.377, SD: 2.001;
hard: 2 8.063, SD: 2.070). Table 10 presents a summary of
descriptive statistics for comprehension scores of unfamiliar and

familiar topic, and for easy and hard versions.

TABLE 10
DESCRIPTIVE STATISTICS FOR COMPREHENSION SCORES

 

 

 

 

 

 

 

Variablfe bcflrnd N Mean Median StDev SEMean
cmpmhns" 0 62 7.649 7.000 2.053 0.261
1 62 6.742 9.000 1.690 0.240
vcbdiff
0 63 8.063 8.000 2.070 0.261
1 61 8.377 9.000 2.001 0.256

 

 

 

 

Code: Cmprehnsn: comprehension
Bckgrnd: prior knowledge; 0: unfamiliar topic; 1: familiar
topic.
Vcbdiff: vocabulary difficulty; 0: hard version/unknown
vocabulary; : easy version/known vocabulary.

__———‘-———_—————_—————————.__—_——————_—————-————_————————

86
The frequency distribution of comprehension scores for familiar and

unfamiliar texts and for hard and easy versions is displayed
graphically in Figures 13, 14, 15, and 16 in Appendix B..

Descriptive statistics were also used to look at the central
tendency and dispersion of the scores within each measure used,
recall (main clauses and supporting clauses) and comprehension
(explicit, Implicit and scriptically implicit questions). Means and
standard deviations for central clauses and supporting clauses are
shown in Table 11.

TABLE 11
EFFECT OF BACKGROUND ON CENTRAL AND SUPPORTING CLAUSES

 

 

 

Variable bckgrnd N Mean Median StDev SEMean
CtrI-Cl 0 62 1.645 1.000 1.392 0.177
1 62 1.613 2.000 1.265 0.163

 

 

SprtrLg-Ci bckgrnd

 

0 62 1.677 2.000 1.142 0.145

 

1 62 2.726 3.000 1.357 0.172

 

Code: Ctrl-Cl: central clauses; Sprtng-Cl: supporting clauses.
Bckgrnd: prior knowledge; 0: unfamiliar topic; 1: familiar
topic.

 

 

At this early stage of the analysis, there were already

indications that prior knowledge affects different aspects of the

87
recall measures. The means were not noticeably different for the

recall of central clauses for the students who read the unfamiliar
text and those who read the familiar topic (Central clauses,
unfamiliar topic: X: 1.645, SD: 1.392; familiar topic: X:1.613, SD:
1.285). Background, however, seemed to have certain effects on the
recall of supporting clauses (Supporting clauses, unfamiliar topic:
X: 1.677, SD: 1.142; familiar topic: X: 2.726, SD: 1.357). Figures 17,
18, 19, and 20 In Appendix B present the histograms for the recall of
central and supporting ideas of unfamiliar and unfamiliar topics.
Vocabulary difficulty also seemed to affect different aspects of
the recall measure. This first analysis of the effect of vocabulary on
the recall of central and supporting clauses seemed to indicate that
vocabulary difficulty affects the recall of both types of clauses. The
level of vocabulary difficulty had some effect on the recall of
central clauses. The mean of the recall of central clauses was .50
higher for those participants that read the easy versions of each
passage (Central clauses, hard vocabulary: X: 1.381, SD: 1.038; easy
vocabulary: X: 1.885, SD: 1.55). The effect of the level of vocabulary
difficulty had an even stronger effect on the number of supporting
ideas recalled than on the number of central clauses (Supporting
clauses, hard vocabulary: X: 1.905, SD: 1.422; easy vocabulary: X:
2.508, SD: 1.22). Table 12 shows the descriptive statistics for the

effect of vocabulary on the recall of central and supporting clauses.

88

 

 

 

 

 

 

EFFECT OF VOCABULARY ONTélgN-TERAIZ. AND SUPPORTING CLAUSES
Variable vcbdiff N Mean Median StDev SEMean
CtrI-CI 0 63 1.381 1.000 1.038 0.131

1 61 1.885 2.000 1.550 0.198
Sprtqu-Cl vcbdiff

0 63 1.905 2.000 1.422 0.179

1 61 2.508 3.000 1.220 0.156

 

 

_-—C-0897—617l:5l;_69ntrgi-_clauses; S—prthZ-Cl: sITp-porting CIaITS—eST——_
Vcbdiff: vocabulary difficulty; 0: hard version/unknown
vocabulary; 1: easy version/known vocabulary.

 

 

 

 

 

Histograms were used to display graphically the frequency
distribution for recall of central and supporting ideas of easy and
hard versions. Figures 21, 22, 23 and 24 In Appendix B show the
frequency distribution of recall scores for hard and easy versions.

Descriptive statistics were also used to look within the
comprehension measure-~explicit, implicit, and scriptically implicit
questions. The results of descriptive statistics for the

comprehension measure are summarized in Tables 13 and 14.

 

 

TABLE 13
EFFECT OF BACKGROUND ON COMPREHENSION MEASURE
Variable Bckgrrid N Mean Median StDev SEMean
Expl-O 0 62 3.00 3.000 0.887 0.113

 

1 62 2.935 3.000 1.143 0.145

89

TABLE 13 (cont'd)

 

 

 

 

ImpI-Q 0 62 2.355 2.000 0.057 0.134
1 62 2.081 2.000 0.929 0.118
ScrptI-Q 0 62 2.339 2.000 0.974 0.124
1 62 3.725 4.000 0.484 0.061

 

——-—--—————————-—_-—_-——————————_-—_————-—-_—-————_—————_——

Code: Bckgrnd: background; 0: unfamiliar; 1: familiar
Epr-O: explicit questions
lmpl-O: implicit questions
Scrptl-Q: scriptically implicit questions

 

 

 

TABLE 14
. EFFECT OF VOCABULARY ON COMPREHENSION MEASURE

 

 

 

 

 

 

 

Variable Vcbdiff N Mean Median StDev SEMean
Epr-Q 0 63 2.810 3.000 1.105 0.139
1 61 3.131 3.000 0.903 0.116
lmpI-O 0 63 2.190 2.000 0.965 0.122
1 61 2.246 2.000 1.043 0.134
Scrptl-Q 0 63 3.063 3.000 1.076 0.136
1 61 3.000 4.000 1.000 0.128

 

 

Code: Vcbdiff: vocabulary difficulty; 0: hard; 1: easy

Expl-Q: explicit questions
lmplco: implicit questions

Scrptl-Q: scriptically

implicit questions

90
Two aspects of these tables are noteworthy. First, background did

not have a strong effect on answering explicit questions. In fact the
mean score was slightly higher for the students who read the ‘
unfamiliar topic compared to the mean score for those participants
who read the familiar topic (unfamiliar: X:3.0, SD: 0.887; familiar:
X: 2.93, SD: 1.143). Students' answers to textually implicit
questions present a similar pattern. The mean score was slightly
higher for the students who read the unfamiliar topic (unfamiliar:
X:2.355, SD: 0.057; familiar: X: 2.081, SD: 0.929). As was expected,
topic familiarity was significant when the participants answered
questions that were scriptically implicit (unfamiliar, X: 2.33,

SD: 0.974; familiar: X: 3.72, SD: 0.484). Second, vocabulary
difficulty affected students' answers to explicit questions. The
mean score was slightly higher for those students who read the easy
version (X: 3.131, SD: 0.90) than the mean for the participants who
read the hard version (X: 2.81, SD: 1.105). When answering textually
implicit questions, the mean score was slightly higher for those
participants who read the easy version (easy: X:2.246, SD: 1.043;
hard: 52: 2.19, so: 0.965).

Figures 25 to 36, in Appendix B display graphically the frequency
distribution of comprehension scores (explicit, Implicit, and
scriptically implicit questions) for familiar and unfamiliar topics,
and for hard and easy versions. The observations of the descriptive
statistics indicated that a multivariate analysis of variance might

be appropriate to use. for this study.

91
Multivariate Analysis of Covariance. The data were then

subjected to a multivariate analysis of covariance (MANCOVA, using
MINITAB 10.0), using the number of units recalled (main clauses and
supporting clauses) and the multiple-choice scores (explicit,
implicit, scriptically implicit questions) as dependent variables.
Topic familiarity, vocabulary difficulty, gender, and whether
participants had visited a Spanish-speaking country were used as
nominal independent variables. Participants‘ age, years that they had
been studying Spanish, length of stay in the Spanish-speaking
country (for those students who had spent sometime overseas),
whether they had contact with the language outside the classroom,
and whether they had Spanish-speaking friends were used as
covariates.

Results indicated that age, gender, years of language study,
interaction with Spanish-speaking friends/relatives, visit to a
Spanish-speaking country, length of the stay in that country, and
contact with the language outside the classroom did not have
statistically significant overall effect on the dependent variables.
The overall effect of topic familiarity was statistically significant
F: 32.186 (5, 109), p : 0.000. The overall effect of vocabulary
difficulty approached statistical significance F : 1.997 (5, 109), p:
.085. At this early stage of data analysis, there were already
indications that the interaction between background knowledge and
vocabulary was not statistically significant F: .804 (5, 109), p:
.549. Tables 15, 16 and 17 present the results of the analysis for

background knowledge, vocabulary difficulty, and interaction

92
between background knowledge and vocabulary; the complete

analysis is Included In Appendix C.

 

 

TABLE 15
MULTIVARIATE ANALYSIS OF COVARIANCEFOR BACKGROUND

KNOWLEDGE ’
naraeonarsaa Tahiti}ER&""’;"1”“E;"{§ "‘5; “53’s "
CRITERION TEST STATISTIC F OF P
Wilk's 0.40381 32.186 (5, 109) 0.000
Lawley-Hotelling 0.47640 32.186 (5. 109) 0.000
Pillai's 0.59619 32.186 (5, 109) 0.000
Roy's 1.47640

TABLE 16

MULTIVARIATE ANALYSIS OF COVARIANCE FOR VOCABULARY

 

 

 

 

MANCOVA for-vocabulary s: 1 m: 1.5 n: 53.5
CRITERION TEST STATISTIC F DF P
Wilk's 0.91606 1.997 (5, 109) 0.085
Lawley-Hotelling 0.09163 1.997 (5. 109) 0.085
Pillai's 0.08394 1.997 (5, 116) 0.065

Roy's 0.09163

#——-*———_—-_-————_——_-———nc———————___—_-——_———————————————.

93

TABLE 17
MULTIVARIATE ANALYSIS OF COVARIANCE FOR INTERACTION BETWEEN
BACKGROUND AND VOCABULARY '

 

 

*——--———-

MANCOV; IOI’ IOpIC fakifiarity _____________
and vocabulary interaction 5: 1 m: 1.5 n: 535

 

 

—-c——— —--—————————————————————-—————__———

CRITERION TEST STATISTIC F OF P
Wilk's 0.96444 0.804 (5, 116) 0.549
Lawley-Hoteliing 0.03687 0.804 (5. 116) 0.549
Pillai's 0.03556 0.804 (5, 116) 0.549
Roy's 0.03687

————_—-———————_———_——-——-—————-—_——————————_———_———————

The first statistical model was trimmed and a separate
multivariate analysis of variance (MANOVA) was performed using
the recall measure and the multiple-choice scores as dependent
variables, and background knowledge, vocabulary, and the interaction
between background knowledge and vocabulary as independent
variables. Even though the first results already Indicated that the
compensatory interaction between prior knowledge and vocabulary
was not statistically significant, this variable was included in this
second analysis because of its relevance to the present study.

With the alpha level set at .05, the results of this second analysis
Indicated that prior knowledge was significant F: .418 (5, 116),

p : .000. The effect of vocabulary approached significance F: .913
(5, 116), p : .059. Because of the Importance of vocabulary to the

present study, and despite the fact that the effect of vocabulary has

94
a p-value slightly higher that .05, this variable will be examined In

the remainder of this dissertation. This position Is supported by
some statisticians who warn us against adopting in advance any
rigid standard. Guilford and Fruchter (1978) commented: ”Instead of
confining ourselves to a two-choice decision--rejection or
acceptance--we might allow a third possibility, that of suspended
judgment...“

Tables 18, 19, and 20 present the results of the analysis of
background knowledge, vocabulary difficulty, and interaction
between background knowledge and vocabulary after the first model

was trimmed.

 

 

 

TABLE 18
MULTIVARIATE ANALYSIS OF COVARIANCE FOR BACKGROUND
— KNOWLEDGE

MANCOVA—Tor topic familiarity .- S= 1“__m:——175-—-_n=:——5_770_—
CRITERION TEST STATISTIC F DF P
Wilk's 0.41814 32.284 (5, 116) 0.000
Lawley-Hotelling 1.39154 32.284 (5. 116) 0.000
Pillai's 0.58186 32.284 (5, 116) 0.000

Roy's 1.39154

95

TABLE 19
MULTIVARIATE ANALYSIS OF COVARIANCE FOR VOCABULARY

 

 

 

MANCOVA for vocabulary s: 1 m: 1.5 n: 57.0
CRITERION TEST STATISTIC F DF P
Wilk's 0.91352 2.196 (5, 116) 0.059
Lawley-Hotelling 0.09466 2.196 (5. 116) 0.059
Pillai's 0.08648 2.196 (5, 116) 0.059
Roy's 0.09466

TABLE 20

MULTIVARIATE ANALYSIS OF COVARIANCE FOR INTERACTION BETWEEN
BACKGROUND AND VOCABULARY

 

 

MANCOVA for topic familiarity _
and vocabulary interaction & 1 m: 1.5 n: 57.0

 

 

CRITERION TEST STATISTIC F DF P
Wilk's 0.96253 0.903 (5, 116) 0.482
Lawley-Hotelling 0.03893 0.903 (5. 116) 0.462
Pillai's 0.03747 0.903 (5, 116) 0.482
Roy's 0.03893

 

In order to analyze both the recall measure and the comprehension

measure further, an analysis of variance (ANOVA) was used to look

96
at the type of information recalled (central clauses and supporting

clauses) and the different question types (textually explicit,
textually implicit, and scriptically implicit).

The univariate test found that vocabulary was statistically
significant on the recall of central clauses (F: 4.59, p< .05). There
was not significant main effect for either background knowledge or
interaction between prior knowledge and vocabulary on the recall of
central clauses. Table 21 presents the results of the analysis of the

recall of central clauses.

TABLE 21
ANALYSIS OF VARIANCE FOR RECALL OF CENTRAL CLAUSES

 

 

 

 

or: SeqSS F P
topic familiarity 1 0.032 0.04 0.844
vocabulary difficulty 1 7.900 4.59 0.034
interaction between
vocabulary difficulty 1 4.383 2.55 0.113

and topic familiarity

 

For recall of supporting clauses, both prior knowledge (F : 22.40,
p < .05), and vocabulary (F : 7.23, p < .05) showed significant main
effects. The interaction between these two variables was not
statistically significant. Table 22 shows the results of the analysis

of the recall of supporting clauses.

97

TABLE 22
ANALYSIS OF VARIANCE FOR RECALL OF SUPPORTING CLAUSES

 

 

 

 

DF Seq 88 F P
topic familiarity 1 34.073 22.40 0.000
vocabulary difficulty 1 10.664 7.23 0.008
interaction between
“MU"? “WWW 1 4.039 2.74 0.101

and topic familiarity

 

Results of the statistical analysis support the claim that
background knowledge and vocabulary difficulty have statistically
significant effects on the students' recalls. These results confirm
claims from L1 reading research regarding the fact that prior
knowledge and vocabulary affect different aspects of the reading
process. Prior knowledge and vocabulary had affected the recall of
supporting ideas. Vocabulary, however, was the only variable that
affected the recall of central Ideas. The failure to find a
compensatory interaction between background knowledge and
vocabulary confirmed findings from research studies in L1 reading.

For the analysis of comprehension scores, the multiple-choice
scores for answers to textually explicit questions, textually
implicit questions and scriptically Implicit questions were used as
dependent variables. Topic familiarity and vocabulary were used as
Independent variables. Because it was hypothesized that background
knowledge and vocabulary function independently, not interactively,

compensatory interaction was examined again.

98
Results of the ANOVA on the reading comprehension scores

indicate that vocabulary and prior knowledge affect comprehension
In different ways. For the students' answers to textually explicit
questions, vocabulary, (F: 3.12, p : .080) had a stronger effect than
background knowledge (F: .15, p : .69). As in the other analysis, the
interaction between the two variables (vocabulary difficulty and
background knowledge) was not statistically significant. These
results agree somewhat with the findings from L1 reading research.
Using a similar reading comprehension measure, Stahl, Jacobsons,
Davis and Davis (1989) found some effect for vocabulary on their
participants' answers to textually explicit questions. The authors,
however, suggest interpreting these results with caution since they
were not “protected by a significant overall effect" (p. 35). Results
of the, statistical analysis of students' answers to textually explicit

questions are presented in Table 23.

TABLE 23
ANALYSIS OF VARIANCE FOR TEXT UALLY EXPLICIT QUESTIONS

 

 

 

 

DF Seq as F P
topic familiarity 1 0.129 0.15 0.695
vocabulary difficulty 1 3,227 3.12 0.080
interaction between
”will” “WW“? 1 0.516 0.50 0.480

and topic familiarity

 

For the participants' answers to textually implicit questions,

neither topic familiarity, (F: 2.42, p : .122), nor vocabulary,

99
(F: .11, p : .738), nor the interaction between these two variables,

(F: 1.53, p : .219), had statistically significant effects. As was
predicted, the analysis showed a significant effect of topic
familiarity, (F: 99.81, p : .000), on answers to scriptically implicit
questions. Vocabulary effect, (F: .38, p : .538), and interaction
effect, (F: .15, p : .700), were not significant. Results of the
statistical analysis of students' answers to textually implicit and

scriptically implicit questions are presented in Tables 24 and 25.

TABLE 24
ANALYSIS OF VARIANCE FOR TEXTUALLY IMPLICIT QUESTIONS

 

 

 

 

 

 

 

 

 

 

DF SeqSS F P
topic familiarity 1 2.230 2.42 0.122
vocabulary difficulty 1 0.111 0.11 0.738
Interaction between
vocabulary difficulty
and topic “mum 1 1.517 1.53 0.219
TABLE 25
ANALYSIS OF VARIANCE FOR RECALL OF SCRIPTICALLY IMPLICIT
QUESTIONS
DF Seq as F P
ionic familiarity 1 59.645 99.81 0.000
vocabulary difficulty 1 0.229 0.38 0.538
interaction between
"mm“? “WWW 1 0.069 0.15 0.700

and toflc familiarity

 

100

4.2 QUALITATIVE ANALYSIS
Students' written protocols

”La Noche de Brujas." in examining the written protocols of the
familiar topic, I found that many of the students who read the easy
version (known vocabulary) of the reading text mentioned the day
when Halloween is celebrated. Although it would be plausible to
suggest that students did that because the topic was extremely
familiar to them, it should also be noted that there was enough
information present in the text to infer the date. Even readers who
were not familiar with the celebration could have inferred the day
when Halloween is celebrated. Eleven students out of the thirty-one
who read the easy version mentioned the day. It seems reasonable to
claim that this was the result of the students' prior knowledge since
this reading was on a familiar topic. However, only two participants
who read the difficult vocabulary text mentioned the date. Unknown
vocabulary appeared to block ready access to prior knowledge.
Excerpts of recalls from students who read the easy version are
shown in Table 26. In order to protect the participants' anonymity,

each written protocol was given a number.

TABLE 26
SAMPLE SENTENCES FROM THE STUDENTS' WRITTEN RECALL
HALLOWEEN/EASY VERSION

 

 

Student #12: (Rsiloween) is—h—e—I-d—B—n—t-he last day or Ootobe}, ———————

Student #13: People celebrate Halloween on the night of
October 31 st.

101

TABLE 26 (cont'd)

Student #36: This celebration takes place on the last day of
October.

—**——————————————_—--————-————————_—————_-————-————_——~-—

Some students who read the difficult vocabulary text also
included information that was not in the reading text. The
information that these students included in their protocols,
however, was of a different type. They mentioned 'children receiving
candy,‘ 'adults wearing masks to spook the kids,‘ and 'children going
with their parents from door to door.‘ It Is interesting to mention
that none of this information was Implied in the text. This suggests
that these students relied more on their prior knowledge of the topic
than did those who read the easy vocabulary text. Table 27 presents

excerpts of recalls from students who read the hard version.

 

 

TABLE 27
SAMPLE SENT ENCES FROM THE STUDENTS‘ WRITTEN RECALL
HALLOWEEN/HARD VERSION
Student # 9: _.. childreh and th—e—ir parents go door to door... ————————

Student #32: On this night adults wear masks to spook the
kids.

Student #33: (Children) go from to house to house collecting
candy.

 

102
The qualitative analysis of the students' protocols also Indicates

that those students who read the hard vocabulary text made 'extra-
textual' inferences. An extra-textual inference is defined. here as a
judgment made or an Idea that is arrived at for which no information
is presented, implied, or suggested by the facts that are given in the
text. Seven students who read the difficult vocabulary version about
Halloween produced extra-textual inferences. None of the students
who read the easy version of the text produced this type of
Inferences. Samples of the students‘ extra-textual inferences are

presented in Table 28.

TABLE 28
SAMPLE SENTENCES FROM STUDENTS' EXT RA-TEXTUAL INFERENCES

 

Sir—6371? TTT-BBB-EITJr-RSES Th; he}? day, November 1, All

Saint's Day, and the two dafi We more
religious.

Student #23: November 1 was named the day of the Saints.
The reason for this is because the people were

Chrigjan and superstitious.

Student #32: it (Halloween) is a Christian tale.

———-—————_——c———————————_————.-—-———————_-——————————_———_

in examining the protocols, I also noticed that parts of some
written recalls written by participants who read the hard version of
the text made little sense. Although some of the paragraphs
contained a central clause and/or a supporting clause, which
was/were counted as such for the quantitative analysis of the data,

the paragraph as a whole lacked coherence and was unclear and

103
confusing. Student #24 wrote: “Apparently it (Halloween) was a

holiday started to keep evil witches away from villages and
signifying the last day of fall. The past to begin winter-some '
villages thought it helped others thought it didn't.“ Student #52
wrote: 'The adults were dreswd in masquerade, as witches, ghosts
and bad spirits. Because the 1st of November marks the beginning of
winter and bad spirits. They celebrate While the Celts celebrate
their holiday.” Protocols from Student #3 and Student #32 presented
similar problems. Even though the number of protocols that
presented such characteristics was small, It is particularly
important to note that all of these protocols were written by
students who read the hard version of the Halloween text. Because of
the unusual characteristics of these recalls, special attention was
given to the protocols that the same students produced after reading
La Diablada, the unfamiliar text. Students #3, 24, and 32 read the
difficult vocabulary text while Student #52 read the easy
vocabulary text. Aside from the amount of information that these
students remembered or did not remember, their protocols were
clear and organized. None of the recalls written by students who
read the easy vocabulary version of Halloween presented this lack of
coherence and organization.

The qualitative analysis of the data indicated that prior
knowledge did not compensate for vocabulary difficulty, and that it
hindered integration of new information. In addition, unknown words
seemed to affect the type of information that was recalled when

reading a familiar t0pic.

104
"La Diablada.” The qualitative analysis of the written recalls of the

unfamiliar topic indicated that unknown vocabulary affected not
only the amount of information recalled but also the kind of
information that was recalled. The first most noticeable difference
between the protocols written by students who read the easy
version and those written by students who read the hard version was
the number of clauses remembered. Those participants who read the
easy vocabulary version recalled over thirty percent more units than
those who read the hard version. Within the group that read the same
version, there were no major differences in the number of central
clauses and supporting ideas that the students recalled. The
participants who read the easy vocabulary text recalled sixty-three
central Ideas and sixty-five supporting ideas. Those participants
who read the hard version recalled thirty-nine central ideas and
thirty nine supporting ideas.

The analysis also indicated that the main effect of vocabulary
difficulty was on the type of information that the students recalled.
Students who read the easy version reported Information directly as
it was in the text or very similar to what was in the passage. No
students reported any inferences, nor did they add any information
that was not in the reading text. On the other hand, the students who
read the difficult vocabulary text reported Information that was
erroneous. Two students, for example, wrote that La Diablada was a
festival celebrated in Pem; one student wrote that it was held in
Brazil and another said that it was a festival in Argentina.

Another interesting pattern found In recalls from students who

read the hard version was the number of participants who wrote how

105
difficult that passage was for them. The fact that they had to read

the difficult version and that they could not rely on their prior
knowledge must have made the task hard for them. Samples of‘the
students' protocols of the unfamiliar topic are presented in Table
29.

TABLE 29
SAMPLE SENTENCES FROM STUDENTS' PROTOCOLS OF THE UNFAMILJAR
TOPIC

Student # 17: “I didn't really understand this passage. It was
something about South American festivals."

Student #18: "i don't know. I don't recall how the party
originated."

Student #38: "I think that the story talked a little bit about
the people who involved, but i don't really
remember. I forgot a lot of the story.“

 

4.3 CHAPTER SUMMARY

The results of this study confirm findings of past research
indicating that prior knowledge and vocabulary difficulty affect the
process of reading. These two aspects, however, do not appear to
interact, and they appear to affect the process of reading
differently. These findings are consistent with earlier L1 reading
research (Freebody and Anderson, 1983b; Stahl and Jacobson, 1986;
Stahl, Jacobson, Davis and Davis, 1989) but in conflict with
Stanovich's interactive hypothesis. The interaction between prior

knowledge and vocabulary was not found to have a primary effect on

106
the recall or comprehension measures used in this study. Vocabulary

was a high predictor for the recall of central ideas and supporting
ideas. Prior knowledge was statistically significant on the recall of
supporting ideas.

On the comprehension measure, vocabulary was the strongest
predictor for the students' answers to textually explicit questions.
Neither prior knowledge nor the interaction between background and
vocabulary were found to have main effects. These findings are also
consistent with the results of Stahl et al.'s research (1986,1989),
which showed that vocabulary had some effect on answers to
textually explicit question in L1 reading.

None of the independent variables had any effect on the
participants' answers to textually implicit questions. These results
also confirmed L1 reading research. On the other hand, the effect of
prior knowledge on the students' answers to scriptically implicit
questions was very significant, as was predicted. The Interaction
between background and vocabulary did not have main effects on the
students' answers to textually implicit questions or to scriptically
implicit questions.

The qualitative analysis of the recall protocols suggests that
prior knowledge and vocabulary affect reading comprehension, but in
different ways. While topic familiarity affected the amount of
information recalled, vocabulary affected the type of information
the participants wrote in their protocols.

In chapter five, the results of the qualitative and quantitative

analyses of the data will be discussed in more depth. I will also

107
refer to the implications of these results for further research and

for the teaching of foreign languages.

CHAPTER FIVE

SUMMARY AND CONCLUSIONS

The main focus of this study was to investigate the effect of

prior knowledge and vocabulary difficulty on the process of reading

in a foreign language. In addition, the possible interaction between

background knowledge and vocabulary was examined. This final

chapter will provide a summary of the research, and discuss the

results and their implications for research in L2 reading and for

foreign language teaching.

The following research hypotheses were formulated regarding the

effect of prior knowledge, vocabulary difficulty, and the interaction

between prior knowledge and vocabulary on recall and comprehension

in L2 reading:

1.

L2 readers who read about a familiar topic will perform
better on the recall of central units than those who read
about an unfamiliar topic.

L2 readers who read about a familiar topic will perform
better on the recall of supporting units than those who read
about an unfamiliar topic.

L2 readers who read an easy vocabulary text will perform
better on the recall of central units than those who read a
difficult vocabulary version.

L2 readers who read an easy vocabulary text will perform

better on the recall of supporting units than those who read a
difficult vocabulary version.

108

10.

11.

12.

13.

14.

15.

109

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity in their recall of central units

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity in their recall of supporting units.

L2 readers who read about a familiar topic will not perform
better on answering textually explicit questions than those
who read about an unfamiliar topic.

L2 readers who read an easy vocabulary text will perform
better on answering textually explicit questions than those
who read a difficult vocabulary version.

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity to answer textually explicit questions.

L2 readers who read about a familiar topic will not perform
better on answering textually implicit questions than those
who read about an unfamiliar topic.

L2 readers who read an easy vocabulary text will not perform
better on answering textually implicit questions than those
who read a difficult vocabulary version.

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary
with topic familiarity to answer textually implicit questions.

L2 readers who read about a familiar topic will perform better
on answering scriptically implicit questions than those who
read about an unfamiliar topic.

L2 readers who read an easy vocabulary text will not perform
better on answering scriptically implicit questions than those
who read a difficult vocabulary version.

L2 readers who read a difficult vocabulary text about a
familiar topic will not compensate for unknown vocabulary

110

with topic familiarity to answer scriptically implicit
quesfions

5.1 SUMMARY OF RESEARCH

One hundred and five students participated in the study. From the
original number of participants, the responses of forty-three
students were not considered In this study because of at least one of
the following reasons: 1) Spanish was the language used at home, 2)
they had been studying Spanish as a foreign language for less than
four years, 3) they were not present in all of the classes when the
study was explained and the data collected. Data were collected on
three consecutive days during the students' Spanish class.

Each student read two passages. One of the passages was about a
familiar topic (Halloween) and the other about an unfamiliar topic
(La Diablada). The participants were randomly given a difficult
(unknown) vocabulary version of the passage or an easy (known)
vocabulary version to read. After the participants finished reading,
they gave the reading texts to the researcher and received a blank
page. They were then instructed to write as much information as
they recalled from the text. Following the free recall, the students
answered twelve multiple-choice questions to assess their
comprehension. The procedure used to write the questions was
modeled on the procedure used by Stahl, Jacobson, Davis and Davis
(1989) in their study on the process of reading in the first language.
According to their coding scheme, four of the questions were
textually explicit; four were textually implicit; and four were

scriptically implicit.

1 1 1
A question whose answer was explicitly stated in the text was

defined as a textually explicit question. A textually implicit
question was defined as one whose answer had to be inferred from
the text. A scriptically implicit question was defined as one whose
answer had to be inferred by the reader on the basis of his/her prior
knowledge; this definition was also used in Pearson and Johnson
(1978), Stahl and Jacobson (1986), and Stahl, Jacobson, Davis, and
Davis (1989). All the questions and possible answers were presented
in English in order to avoid comprehension problems.

The data in this study were subjected to three different types of
statistical analysis. First, a multivariate analysis of covariance
was performed on the students' free recalls and their answers to the
multiple-choice comprehension questions. Second, the model was
trimmed and a multivariate analysis of variance was performed on
the recall and the comprehension measures. in order to test all the
hypotheses of the study, the recall measure (recall of central and
supporting units) and the comprehension scores (answers to
textually explicit questions, textually implicit questions and
scriptically implicit questions) were subjected to various analysis
of variances.

To analyze the students' free recalls further, their written
protocols were individually examined. This qualitative analysis
focused on the type of information included In the students' written

protocols.

1 12
5. 2 CONCLUSIONS

The findings of this study confirmed results from L1 and L2
research indicating that prior knowledge and vocabulary difficulty
affect comprehension processes and recall from text. However,
these two factors appeared to affect different aspects of the
process of reading.

Prior research and common sense suggest that background
knowledge facilitates recall. However, in analyzing how prior
knowledge affected recall of central ideas, this research does not
support hypotheses 1. The findings here suggest that, in fact, topic
familiarity did not facilitate the recall of central units. The means
for the recall of central units were not significantly higher for
those students who read the topic about Halloween. Background
knowledge did, however, have a stronger effect on the recall of
supporting ideas. These results supported hypotheses 2. One of the
reasons for this may be that the topic was so familiar to the
students that they relied too much on their prior knowledge. Their
schemata not only did not help them, but actually interfered with
the Integration of the new information. Most of the ideas that were
considered central in the reading text explained the origin of
Halloween. Some of the clauses that were classified as central units
read: 'This popular celebration actually started as a Celtic festival,‘
and 'The last day in October was dedicated to The Lord of the Dead.‘
It is quite likely that these ideas were totally new to the students.
On the other hand, most supporting ideas presented information that
was known by the students, such as 'La Noche de Brujas' is the

favorite holiday of many children in North America,’ and 'The

1 13
children and sometimes adults too, wear masks...‘ This may explain,

in part, the different effect that background knowledge had on the
recall of central and supporting units. In this study, prior knowledge
did not seem to help the participants encode new information from
the text.

Schema theory claims that our experience results in knowledge
that is stored in memory, in slots or frameworks usually called
schemata. A schema will help the reader interpret a new situation
and integrate new information. Alba and Hasher (1983) identified
five central cognitive processes in schema theory. Four of these
processes were identified as encoding processes (selection,
abstraction, interpretation, and integration) and one as a retrieval
process. In this study, prior knowledge did not help the students
recall new information that had been categorized as central units.
The students' inability to recall central units suggests that the
interconnection of the processes within the encoding system, and
between the encoding system and the retrieval system may not be as
automatic as Is often assumed.

These findings appear to be easier to explain under dual coding
theory. As it was discussed in Chapter 1 (p. 31), this theory suggests
that cognition consists of two mental subsystems that specialize in
the representation and processing of information. These two
subsystems, verbal and nonverbal, have the capacity to operate in an
integrated manner. This Integration facilitates the processing of
information In reading. The verbal subsystem specializes In
language. It involves the processing of linguistic stimuli. The

nonverbal systems involves the generation of mental Images while

114
processing visual, auditory and affective information. The

information within each subsystem is Interconnected, and there are
referential interconnections between the two subsystems. Incoming
information is assumed to be stored under two mental
representations In two separate subsystems. Because different
variables may affect the information within one subsystem and the
referential interconnection between the two subsystems, supporters
of dual coding theory explain that some variations can occur at the
time of the retrieval. Results of this study indicate that unknown
vocabulary could be one of the variables that affects the
representational connections within each system. This results in the
readers' difficulty to use their background knowledge to integrate
new information. Difficult vocabulary may also affect the

referential connections between the two subsystems limiting the
readers' ability to construct a coherent recall at the time of
information retrieval.

According to the results of this study, difficult vocabulary
affected the recall of both central units and supporting units. These
findings show hypotheses 3 and 4 to be true for the current study.
Students who read the easy vocabulary text performed better in the
recall of central and supporting clauses than those students who
read the difficult vocabulary text.

Prior research has provided robust evidence that both background
knowledge and vocabulary affect the process of reading. Since a
person who is very familiar with a particular topic generally knows
words related to that topic, it is believed that these two factors are

related to each other. Research in L1 reading investigated the

115
interactive-compensatory hypotheses presented by some

researchers such as Rumelhart (1977) and Stanovich (1980, 1984).
These research studies, however, failed to demonstrate. an overall
significant interaction between these two factors (Freebody and
Anderson, 1983b; Stahl and Jacobson, 1986; Stahl, Jacobson, Davis
and Davis, 1989). A text that contains difficult (unknown) words
will hamper recall even if the reading is about a familiar topic. The
results from this study were consistent with findings In L1
research. The participants could not use their high-level knowledge
(background knowledge) source to compensate for vocabulary
difficulty in the recall of central and supporting units. These
findings support hypotheses 5 and 6 of this study.

The qualitative analysis of the students' protocols also suggested
that background knowledge and vocabulary affected reading
differently. Some students included in their recall protocols
information that was not in the text they read about the familiar
topic. The students' familiarity with the topic was the grounds for
all that 'extra' information reported. However, two trends were
found upon examination of the information that these students
included in their reports. Vocabulary appeared to affect the type of
information they recalled. On the one hand, students who read the
easy vocabulary version mentioned when Halloween is celebrated. it
can be argued that this occurred because of the students' familiarity
with the topic. But it could have also been inferred from the text.
Good readers could have inferred when Halloween is celebrated even
if they were not so familiar with the celebration. On the other hand,

it is important to note that the students who read the difficult

1 1 6
vocabulary text added information that could have never been

inferred from the reading.

Such a pattern suggests that vocabulary interfered with recall of
information. It is possible that the students' effort to extract
meaning demanded too much mental energy because of vocabulary
difficulty. This strain on the participants' attention capacity did not
leave enough mental energy for the students to process the rest of
the information. Then, when the students were asked to write down
as much Information as they recalled, prior knowledge was the only
cognitive resource they could rely on. This resulted in students
reporting information that was neither in the text nor could have
been Inferred.

Difficult vocabulary also appeared to affect the clarity and
cohesion of the students' recall protocols. Parts of some protocols
were disorganized and incoherent. This indicated that participants
recalled parts of the text, but difficult words hindered their ability
to process the information and recall it in a logical and coherent
way. Stahl et al. (1989) also found that when their students read a
difficult vocabulary text, the ideas in their protocols were
disorganized and lacked coherence.

As was hypothesized, only vocabulary showed significant effect
on answers to textually explicit questions. Although it would be
reasonable to think that prior knowledge facilitates answers to
textually explicit questions, findings of this study did not come as a
surprise since earlier L1 research found similar results (Stahl et al,
1989). No main effect for prior knowledge and vocabulary

interaction was found. This failure to find a compensatory

117
interaction effect on answers to textually explicit questions

strengthens the claim that prior knowledge and difficulty of
vocabulary function independently. These results support hypotheses
7, 8, and 9.

According to results of this research, neither background
knowledge nor vocabulary appeared to have any effect on answers to
textually implicit questions. Anderson and Pearson (1984) suggested
that one of the ways in which prior knowledge facilitates
comprehension is by enabling the reader to make inferences about
their reading. Results from this study did not support this claim.
Participants were expected to pull together information from
different sentences in order to answer textually Implicit questions.
The following is an example of a textually implicit question.

The . students read:

(Los Celtas) Cretan que en esta época del afio las
brujas, los malos espiritus y hasta el mismo dlablo
vlsltaba los pueblos y villas. La gente entonces
encendia grandee fuegos que mantenlan encendldos
durante toda la noche, pues se crefa que de este modo
se asustaban esas malas Influenclas. Otros puebios en
Europa tamblén pensaban que durante el ultimo dia de
octubre Ios malos espiritus andaban por el mundo.

The Celts believed that it was at this time of the year when
witches, evil spirits, and the Devil, himself, visited towns and
villages. People lit fires that kept burning all night to ward off
evil spirits.

People in other parts of Europe also believed that evil spirits
roamed the world during the last day In October.

The textually implicit question that the students had to answer
read:

Why might other groups In Europe have llt tires on
October 31?

118

(a) Because the season of darkness started.
(b) To ward off evil spirits.
(c) To warship the Sun God.

(d) Because they believes in Samhain.

Item 'b' was the correct answer. Readers failed to infer that
because other people in Europe also believed that evil spirits visited
the world on October 31, they might have lit fires to ward off those
bad spirits. It is not clearly known why the students' prior
knowledge did not facilitate the answer to textually implicit
questions. The reason may lie in the fact that the information was
so new to them that their schema did not encode and store such
information.

The effect of the interaction between prior knowledge and
vocabulary was not significant on answers to textually implicit
questions. These results support hypotheses 10, 11, and 12 of this
research.

As was hypothesized, only background knowledge showed
significant effects on answers to scriptically implicit questions.
Students performed better in answering scriptically implicit
questions when they read about the familiar topic than when they
read the unfamiliar topic. The only cognitive source students could
use to answer this type of questions was their prior knowledge.
Therefore, when the participants had to answer questions about La
Diablada, they lacked the schemata necessary to answer those
questions correctly. Following are examples of scriptically implicit
questions that the students had to answer about 'La Noche de Brujas'
and 'La Diablada':

119

What colors are usually associated with 'La Noche de
Brujas?

(a) Red and black.

(b) Black and orange.

(c) Orange and green.

(d) Green and red.

What metal was extracted In large quantities from an
area near Oruro?

(a) Gold

(b) Silver

(c) Copper

(d) Zinc

Vocabulary did not affect the students' answers to scriptically
implicit questions. Most of the students answered these questions
correctly when they read the passage about Halloween. Only two
students out of sixty-two answered the example above Incorrectly.
One of.those students had read the easy version and the other the
hard version, and they both chose item 'a' as the correct answer. The
fact that the color red is often used for items associated with
Halloween might have affected the way these two students
answered this particular question. Also, option 'a' was the one next
to the correct answers, so the possibility that the students marked
'a' by mistake can not be disregarded. On the other hand, when the
students read aboutthe unfamiliar topic, La Diablada, they did not
have the schemata to help them answer scriptically implicit
questions. For example, only twelve students correctly answered the
question presented above about 'La Diablada.“ Most students chose
item 'a. gold' as the correct answer, probably because of the high
value that is usually attached to gold. Thirteen participants chose

'oopper,‘ and two students chose 'zinc.‘

120
As in the other analysis, no significant interaction was found

between prior knowledge and vocabulary. These results confirm
hypotheses 13, 14, and 15 of this study.

Before closing this section on conclusions, it is important to
briefly discuss the students' reactions in the classroom after they
read the different passages and versions. When the students read the
passage about Halloween, the comments that were heard in the
classroom were not directly related to any aspects of the text. The
comments were mostly humorous because they read about Halloween
in May. Some of the students' comments were: “Are we celebrating
Halloween earlier this year, or what?“ “Somebody should have told
us that Halloween has been changed..." _‘

The students‘ comments after they read La Diablada were about
the reading text. They complained that they had been asked to read
about La Diablada, something they did not even know existed. They
said that there were “too many“ things that they "did not understand"
and, of course, “could not include“ in the recalls because “they did
not remember much.“ To these comments, students who had read the
hard version added that there were "too many words" that they did
not know. One student who had read the easy version said ”it was not
the words. I knew all the words... I think, it was the passage... the
whole passage was hard.”

It is interesting to note that the students mentioned the inclusion
of unknown words when they read La Diablada. They did not make any
reference to the difficult vocabulary when they read La Noche de

Brujas. The cognitive cost of having to read about a completely

121
unfamiliar topic made them become aware of all the unknown words

that were in the passage.

5.3 IMPLICATIONS FOR L2 READING RESEARCH

Before discussing the implications of the present study for
further research In the area of reading In a foreign language, it
would be reasonable to examine the limitations of the study. Three
words in the unfamiliar topic text which had been classified as 'easy
words' (known) were actually identified as unknown by the students.
Although the number is really small, it is unknown whether those
words could have affected the analysis of the effect of vocabulary
difficulty.

The results of the present study pose questions, which should be
addressed In future investigations. Further research is needed to
know whether results from this study can be generalized to other
age groups. The students who participated in this study were all
teenagers. Although the analysis of the data Indicated that age did
not have any effect on the results, it would be interesting to find
out if the same results would be obtained with children or with
adults.

Research in L2 reading has provided evidence suggesting that
orthographic systems affect word recognition and reading. In a very
widely cited study, Scriber and Cole (1981) conducted a series of
cognitive tasks with four groups of Vai adults, an African tribe. One
group included Arabic literates; another group included English
literates; another group Included Vai Iiterates, and the fourth group

included illiterate adults. Scribe and Cole claimed that their

122
subjects performed better on cognitive tasks that required cognitive

operations that were similar to the orthographic system that they
usually read. In more recent studies Koda (1987, 1992) and Shimron
and Sivan (1994) investigated the effect of orthography on the
process of reading. Important results would come from research
with a design similar to the present study but with participants who
use different orthographic systems.

Replacing approximately every sixth content word by a difficult
synonym appeared to affect the recall of central and supporting
ideas. It is generally assumed that readers can skip words and still
continue to comprehend. Further research is needed to investigate
what ratio of difficult words affects recall of central and
supporting ideas.

Results from this study suggest that prior knowledge and
vocabulary affect different aspects of reading. More research using
different text structures and vocabulary is needed to better
understand how prior knowledge and vocabulary affect
comprehension, and to clearly identify which aspects these two

variables affect.

5.4 IMPLICATIONS FOR L2 TEACHING

For many years reading teachers have been using prereading
activities to either activate the students' background knowledge or
to provide relevant information to facilitate reading comprehension.
The results of the present study support the claim that prior
knowledge affects information recall and comprehension. However,

it may be imprudent to believe that topic familiarity will help

123
readers cope with vocabulary difficulty. Foreign language teachers

should be aware that unknown words may be far more troublesome
than has been usually thought.

One of the main reasons why people read in a foreign language is
to learn, to acquire knowledge. Most second/foreign language
Ieamers see the language as a tool. There may be L2 readers who
read for pleasure, but the vast majority do it to get information that
will help them improve their work quality, their efficiency at work,
or simply help them pass a test. Results of this study indicate that
vocabulary plays a very important role in the recall of main and
supporting ideas. By developing activities that help L2 language
learners enlarge their vocabulary, language teachers will help them
become more efficient and effective readers and more successful
language users.

An excellent way to help second language readers enlarge their
vocabulary is to use contextually rich materials. The regular use of
semantically rich texts allows language Ieamers to use context to
derive meaning of unknown words. Klein (1988) refers to a personal
experience related to this point. His stepdaughter insisted on
reading to him some news from the newspaper. The article was
about an interview to a police officer. The article read: '... "I was
speeding because I thought my plant was moribund." She told the
officer that the plant did not look well, was very dry, the leaves
were wilted, and it was very expensive.‘ When the author asked the
girl whether she knew what the word moribund meant, the girl
responded that "it was something that was either very sick or dead!"

Klein states that this is not the normal use of the term moribund,

124
the girl, however, was able to derive the accurate meaning using the

semantically rich context in which the word was embedded. Readers
should became aware that they can depend on contextual information
in the text to generate meaning for new word when reading in a
foreign language.

Since students cannot be taught every word they will encounter
as they read, some specialists suggest the use of various reading
texts on the same topic. This helps language learners not only to
enlarge their vocabulary but also to develop the ability to make
inferences as they become more familiar with the topic and the
vocabulary.

Students should understand the advantages of using their prior
knowledge to obtain a better comprehension of the text. Language
learners are not always aware that personal experience is very
important In reading comprehension. Students should be encouraged
to think about how their reading relates to what they already know,
and to think about What else they know about the topic. Language
teachers should provide their students with reading activities and
work with the learners to help them develop the ability to use their
background knowledge when they read in a foreign language.

Unknown vocabulary presents a difficult problem to L2 reading
comprehension. Not only do difficult lexical items hinder the recall
of Important information, but also appear to affect the coherence
and organization of their recalls. Foreign language teachers should
continue developing prereading activities to help their students
activate their background knowledge. Some activities should also

focus on word meaning since there seems to be strong evidence that

125
indicates that vocabulary knowledge is central to succeed in certain

reading comprehension tasks.

APPENDIX A
STUDENT MATERIALS

APPENDIX A
STUDENT MATERIALS

W

The research study being done by Ms. Susana B. Tuero is part of a
project for a doctoral dissertation on the process of reading in a
foreign language.

The activities that will be done during the time the data are
collected will not bring any changes into the classroom or in the
way classes are usually conducted. It must also be noted that your
perforrnanoe on the activities will not affect your grades in any way.
After the data are analyzed, the general results will be available to
the language teachers and the students that participated in the
study.

' The research has been explained to me and I understand the basic
purposes and procedures. I freely consent to participate In the study
and understand that at any time I may withdraw without penalty.

I understand that all the results will be treated with strict
confidence and that all subjects will remain anonymous; upon
request, results will be made available to me as long as the

anonymity of other subjects can be maintained.

——-———————_—-————————__-—o_—c—

————————_———————-—_——————.—

127

TUDENT ESTI NNAIFIE

1. Name:

 

2. Age:

 

3. Gender: F M

—-—. -—_—

4. Language spoken at home:

 

5. How long have you been studying Spanish at school?

 

6. Have you ever visited any Spanish speaking countries?
Yes_ No

—— ——-

If your answer is ‘yes,' when, where, and how long?

 

-————_——————_--————_—-————_———————_————--———-—————*—_-—

7. Have you ever lived in a Spanish speaking country?
Yes No

-—_ ————

If your answer is 'yes,' where and how long?

 

__—_————————————-———————_———————————.—.—-——_————-———*—_—_—

8. Do you have Spanish speaking friends/relatives?
Yes No

If your answer is 'yes,'

do they speak Spanish to you? Yes_ No___
do they speak in Spanish to each other while you are
around? Yes___ No

9. Do you have any contact with the Spanish language outside the
classroom (reading books, watching TV, and/or listening to the
radio in Spanish)? Yes___ No___

10. Are you planning to continue studying Spanish after you graduate

from high school? Yes___ No

——-—-

128

READING PASSAGE-FAMILJAR TOPICIEASY VERSION

W

Una de las fiestas mas esperadas por los ninos en Norteamérica
es la famosa “Noche de ijas." Ese dla por la tardecita, Ios nifios
van de puerta en puerta visitando amigos y vecinos. Los nifios y a
veces también Ios adultos llevan mascaras, algunas muy feas, que
representan personajes de ciencia ficcién, figuras del gobiemo y
muy especialmente brujas, fantasmas y esqueletos.

Esta popular celebracion se inicio en realidad como un festival
Celta. Los Druidas eran Ios maestros y sacerdotes de Ios Celtas,
adoraban el sol y consideraban al fuego un elemento sagrado. El
ultimo dia de octubre lo dedicaban a Samhain, el Sefior de la Muerte.
Este era también el ultimo dia en el calendario celta. En noviembre
empezaba el invierno, la estaclén del frio, la tristeza y la oscuridad.
Por esto es que los Celtas asociaban el primer dia de noviembre con
los muertos. Creian que en esta época del afio las brujas, Ios malos
espiritus y hasta el misrno diablo visitaba Ios pueblos y villas. La
gente entonces encendia grandes fuegos que mantenlan encendidos
durante toda la noche, pues se crela que de este modo se asustaban
esas malas influencias. Otros pueblos en Europa también pensaban
que durante el ultimo dia de octubre los malos espiritus andaban por
el mundo.

En el aho 834, el Papa Gregorio lV declaro el 1 de noviembre como
el dia de todos Ios santos. Y fue durante la dominacién romana que la
celebracién celta se mezclo con la celebracién cristiana. Hoy pocos
son Ios que conocen el origen de la “Noche de Brujas," y como dlcha
celebracién llego hasta nosotros.

129

READING PASSAGE-FAMILIAR TOPIC/HARD VERSION

W

Una de las fiestas mas esperadas por los nlfios en Norteamérica
es la famosa “Noche de Brujas.“ Ese dia por la tardecita, Ios ni‘r'ios
van de puerta en puerta visitando amigos y vecinos. Los nifios y a
veces también Ios adultos llevan mascaras, algunas horripiiantes,
que representan personajes de ciencia ficcién, figuras
gubernamentales y muy especialmente brujas, fantasmas y
esqueletos.

Esta popular celebracién se comenzo en realidad como un festival
Celta. Los Druidas eran Ios gulas y sacerdotes de Ios Celtas,
adoraban el sol y consideraban al fuego un elemento mistlco. El
ultimo dIa de octubre lo dedicaban a Samhain, el Sefior de la Parca
Este era también el ultimo dia en el calendario celta. En noviembre
comenzaba el lnviemo, Ia estacién del frlo, Ia tristeza y la negrura.
Por esto es que los Celtas asociaban el primer dia de noviembre con
los difuntos. Creian que en esta época del afio las bmjas, Ios malos
espiritus y hasta el misrno mandinga visitaba Ios pueblos y villas. La
gente entonces encendia grandes hogueras que mantenlan encendidas
durante toda la noche, pues se crela que de este modo se auyentaban
esas malas influencias. Otros pueblos en Europa también pensaban
que durante el ultimo dia de octubre Ios malos espiritus andaban por
el mundo.

En el ai‘io 834, el Papa Gregorio IV instituyo el 1 de noviembre
como el dia de todos los santos. Y fue durante la dominacién romana
que la celebracién celta se entretejio con la celebracién cristiana.
Hoy pocos son los que conocen el origen de la “Noche de Brujas," y
como dicha celebracién Ilego hasta nosotros.

130

READING PASSAGE-FAMILIAR TOPIC (TRANSLATION)

W

"La Noche de Brujas" is the favorite holiday of many children in
North America. On this day, in the evening, children go from door to
door, visiting friends and neighbors. The children and sometimes
adults too, wear masks, some of them really ugly, that represent
science fiction characters, figures from the government and
specially witches, ghosts, and skeletons.

This popular celebration actually started as a Celtic festival. The
Druids were the guides and priests of the Celts; they worshipped the
sun and considered fire a sacred element. The last day in October
was dedicated to Samhain, The Lord of the Dead. This day was also
the last day on the Celtic calendar. November indicated the beginning
of winter, the season of cold, sadness, and darkness. This was the
reason Why the Celts associated November 1 with the dead. The
Celts believed that it was at this time of the year when witches,
evil spirits, and the Devil, himself, visited towns and villages.
People lit large fires that kept burning all night to ward off evil
spirits. People in other parts in Europe also believed that evil
spirits roamed the world during the last day in October.

In the year 834, Pope Gregory IV established November 1 as All
Saint's Day. It was during the Roman domination that the Celtic
festival and the Christian faith amalgamated. Today few people
know the origin of "La Noche de Brujas," and how this celebration

came 10 US.

131

WRITTEN RECALL-FAMILIAR TOPIC

(Name)

You have just read a short passage about La Noche de Brujas. In

the space provided below, write in English as much information from
the passage as you recall. Please, write down your ideas in the same
order you think they were presented in the text. Use the back of this

paper if necessary.

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
............................................................................................................
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

1 3 2
COMPREHENSION ASSESSMENT -FAMILIAR TOPIC

(A) .............................................

The questions below are about La Noche de Brujas. Read them
carefully and choose one item 'a,‘ 'b,’ 'c,‘ or 'd' that answers each
question correctly.

1. What colors are usually associated with La Noche de Brujas?
(a) Red and black.
(b) Black and orange.
(c) Orange and green.
(d) Green and red.

2. What can be see in most houses on La Noche de Brujas?
(a) Pumpkins and gourds.
(b) Poincettias and mistletoe.
(c) Watermelons and plums.
(d) Pine wreaths.

3. Why did the Celts associate November 1 with darkness?
(a) Because all the fires in their homes were put out that day.
(b) Because they believed that the Lord of Death visited all
towns that day.
(c) Because they believes that evil powers could enter their
homes that day.
(d) Because it was the beginning of winter for them.

4. Who were the Druids?
(a) Soldiers that defended the Celts.
(b) Kings of different Celtic tribes.
(c) Priests and guides for the Celts.
(d) Gods worshipped by the Celts.

5. Why do children go from house to house on La Noche de Brujas?
(a) To sing to their friends and neighbors.
(b) Because friends and neighbours give them candy.
(0) Because they have candy for their friends and neighbors.
(d) Because they sell candy to their friends and neighbors.

 

133

6. What was the last day in October according to the Celtic

10.

11.

12.

calendar?

(a) It was the day when the Celts started harvesting their crops.
(b) It was the day when the Celts began preparing for war.

(c) It was the last day of the year on the Celtic calendar.

(d) It was when the Celts started their sacred fire each year.

Why might other groups in Europe have lit tires on October 31?
(a) Because the season of darkness started.

(b) To ward off evil spirits.

(c) To worship the Sun God.

(d) Because they believed in Samhain.

Why was the fire so important for the Celts?

(a) Because they related it to God.

(b) Because they used it to honor the dead.

(c) Because it brought strength to the feeble folk.
(d) Because it symbolized eternal life.

How did this Celtic celebration come to us?

(a) Because many people believed in witches.

(b) Because the Celtics spreaded their beliefs all over Europe.
(0) Because it mixed with the Christian faith.

(d) Because the Romans adopted the Celtic calendar.

What happened in the year 834?

(a) The Romans conquered the Celtic tribes.
(b) Good people lit their fires in Europe.

(c) The Celts stopped worshipping Samhain.
(d) The Pope established the All Saints‘ Day.

On whom do children play tricks on La Noche de Brujas?
(a) 0n the people who dress up as witches and ghosts.
(b) On people who lit fires that night.

(c) On people who don't have candles by their windows.
(d) On people who don't give them anything.

What was the relation between October 31 and death for the
Celts?

(a) October 31 and death represented the end of something.
(b) On October 31, the Celts lit fires to honor the dead .

(c) Gods that protected the dead were worshipped on October 31.
(d) Celtic Kings were placed in their burial sites on October 31

(B)

1 3 4
COMPREHENSION ASSESSMENT-FAMILIAR TOPIC

The questions below are about La Noche de Brujas. Read them
carefully and choose one item 'a,’ 'b,’ 'c,‘ or 'd' that answers each
question correctly.

1.

Why do children go from house to house on La Noche de Brujas?
(a) To sing to their friends and neighbors.

(b) Because friends and neighbours give them candy.

(c) Because they have candy for their friends and neighbors.
(d) Because they sell candy to their friends and neighbors.

What can be see in most houses on La Noche de Brujas?
(a) Pumpkins and gourds.

(b) Poincettias and mistletoe.

(c) Watermelons and plums.

(d) Pine wreaths.

Why did the Celts associate November 1 with darkness?

(a) Because all the fires in their homes were put out that day.

(b) Because they believed that the Lord of Death visited all
towns that day.

(c) Because they believes that evil powers could enter their
homes that day.

(d) Because It was the beginning of winter for them.

Why was the fire so important for the Celts?

(a) Because they related it to God.

(b) Because they used it to honor the dead.

(0) Because it brought strength to the feeble folk.
(d) Because it symbolized eternal life.

Who were the Druids?

(a) Soldiers that defended the Celts.
(b) Kings of different Celtic tribes.
(c) Priests and guides for the Celts.
(d) Gods worshipped by the Celts.

135

6. What was the relation between October 31 and death for the

10.

11.

12.

Celts?

(a) October 31 and death represented the end of something.

(b) On October 31, the Celts lit fires to honor the dead .. ..
(c) Gods that protected the dead were worshipped on October 31.
(d) Celtic Kings were placed in their burial sites on October 31

What was the last day in October according to the Celtic
calendar?

(a) It was the day when the Celts started harvesting their crops.
(b) It was the day when the Celts began preparing for war.

(c) It was the last day of the year on the Celtic calendar.

(d) It was when the Celts started their sacred fire each year.

Why might other groups in Europe have lit fires on October 31?
(a) Because the season of darkness started.

(b) To ward off evil spirits.

(c) To worship the Sun God.

(d) Because they believed in Samhain.

How did this Celtic celebration come to us?

(a) Because many people believed in witches.

(b) Because the Celtics spreaded their beliefs all over Europe.
(c) Because it mixed with the Christian faith.

(d) Because the Romans adopted the Celtic calendar.

What colors are usually associated with La Noche de Brujas?
(a) Red and black.

(b) Black and orange.

(c) Orange and green.

(d) Green and red.

What happened in the year 834?

(a) The Romans conquered the Celtic tribes.
(b) Good people fit their fires in Europe.

(c) The Celts stopped worshipping Samhain.
(d) The Pope established the All Saints' Day.

On whom do children play tricks on La Noche de Brujas?
(a) On the people who dress up as witches and ghosts.
(b) On people who lit fires that night.

(0) On people who don't have candles by their windows.
(d) On people who don't give them anything.

136

READING PASSAGE-UNFAMILIAR TOPIC/EASY VERSION

W

Cada afio se celebra en Omro, Bolivia, la famosa “Diablada", una de
las fiestas mas interesantes de Sudamérica. La poblacién de Oruro
esta compuesta, en su mayor parte por indios. Todos los afios en
tiempo de Camaval, en febrero 0 marzo, estos indios abandonan sus
peligrosos trabajos en las minas y se ponen a bailar al ritmo de las
guitarras y otros instrumentos tipicos. Los bailarines llevan
mascaras que representan diferentes figuras de la historia y de la
mitologia, asl como también sapos serpientes y especialmente
diablos, centenares de diablos. Muchos de los turistas que llegan a
ver La Diablada no saben del doble significado de la fiesta.

Cuenta Ia leyenda de los Andes que hace mucho tiempo, el antiguo
pueblo andino de los Urus llevo una vida de mucha libertad, pero que
finalmente, conocio el arrepentimiento. Un dia cambio su manera de
vivir y abandono las malas costumbres. Sin embargo, al antiguo
senor de los Urus, un espiritu malo llamado Huarl, no le gusto este
cambio y resolvio vengarse. Trato de vencer a los Urus envlandoles
muchos y diferentes animales para que los eliminaran. Entonces
aparecio Nusta, una princesa Inca, que transformo a algunos
animales en figuras de piedra dura y al resto de los animales que
venian a atacarlos en pequei‘ios granos de arena.

Para muchos de los habitantes de la zona y para la mayor parte de
los indios, Ia Nusca es la Virgen del Socavén, patrona de los mineros
y figura central de La Diablada. El desfile de musicos y bailarines es
seguida con interés por lugarefios y vlsitantes. El origen de la fiesta,
los bailes caracteristicos y los brillantes colores de la vestlmenta
de los participantes le dan a esta fiesta su espiritu unico.

137

READING PASSAGE-UNFAMILIAR TOPIC/HARD VERSION

W

Cada aflo se celebra en Oruro, Bolivia, la famosa "Diablada", una de
las fiestas mas interesantes de Sudamérica. La poblacién de Oruro
esta compuesta, en su mayor parte por quechuas. Todos Ios afios en
tiempo de Camaval, en febrero 0 marzo, estos indios dejan sus
peligrosos trabajos en las minas y se ponen a bailar al ritmo de Ios
charangos y otros instrumentos tipicos. Los bailarines llevan
caretas que representan diferentes figuras de la historia y de la
mitologia, asl como también batracios, serpientes y especialmente
diablos, centenares de diablos. Muchos de los turistas que llegan a
presenciar La Diablada no entienden el doble significado de la
fiesta.

Cuenta la leyenda de los Andes que en tiempos inmemoriales, el
antiguo pueblo andino de Ios Urus llevo una vida disipada, pero que
finalmente,. conocio el arrepentimiento. Un dia cambio su manera de
vivir e hizo a un lado las malas costumbres. Sin embargo, al antiguo
senor de Ios Urus, un espiritu maligno llamado Huari, no le gusto
este cambio y resolvlo vengarse. Trato de subyugar a los Urus
enviandoles muchos y diferentes animales para que Ios eliminaran.
Entonces surglo Nusta, una princesa Inca, que transformo a algunos
animales en esfinges de piedra dura y al resto de Ios animales que
venian a atacarlos en minusculos granos de arena.

Para muchos de los habitantes de la zona y para la mayor parte de
los quechuas, la Nusca es la Virgen del Socavbn, patrona de los
mineros y figura irremplazable de La Diablada. El desfile de muslcos
y bailarines es seguido con interés por lugarehos y forasteros. El
origen de la fiesta, Ios bailes caracteristicos y los brillantes
colores de la vestimenta de los participantes le dan a esta fiesta su
espiritu unico.

1 38
READING PASSAGE-UNFAMILIAR TOPIC (TRANSLATION)

La Diablada ls held each year in Oruro, Bolivia. Most of the
inhabitants in Oruro are Indians. Every year at Carnival time, in
February or March, all the Indians leave their dangerous jobs in the
mines to dance to the music of guitars and other traditional musical
instruments. All the dancers wear masks which look like historical
and mythological figures or like toads, snakes, and devils, hundreds
of devils. They are part of a parade which marches through the
streets of the town. Many of the tourists that come to see La
Diablada do not know the double origin of the festival.

According to the Andean legend, a long time ago the ancient
Andean Urus lead a very lascivious life. As time went by, they
repented and changed their lifestyle. However, Huari, the Lord of the
Urus, did not like this change and sought revenge. Huari sent several
different animals to exterminate the Urus. Then Nusca, an lncan
princess, appeared and turned some of the animals into stone figures
and others into small grains of sand.

For many inhabitants of the region and for most Indians, Nusca is
the Virgin of Socavdn, protector of all the miners and central figure
of La Diablada. The parade of musicians and dancers is avidly
watched by residents and visitors. The origin of the festival, the
traditional dances and the bright colors of the participants'

costumes give this festival its unique spirit.

1 39
WRITTEN RECALL-UNFAMILIAR TOPIC

You have just read a short passage about La Diablada. In the space
provided below, write in English as much information from the
passage as you recall. Please, write down your ideas in the same
order you think they were presented in the text. Use the back of this

paper if necessary.

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
............................................................................................................
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
............................................................................................................
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

140

COMPREHENSION ASSESSMENT-UNFAMILIAR TOPIC

(A) ............................ . ................

The questions below are about La Diablada . Read them carefully
and choose one item 'a,’ 'b,‘ 'c,‘ or ‘d' that answers each question
correctly

1. What is La Diablada?
(a) A group of people
(b) A celebration
(c) A popular dance
(d) A type of music

2. Where do people celebrate La Diablada?
(a) In a Caribbean country.
(b) In a country in South America.
(0) In a region In Mexico.
(d) On some Spanish islands.

3. What is the origin of La Diablada?
(a) It's a mixture of myth and religion.
(b) It's the history of a group of native Indians.
(c) It's completely unknown.
(cl) It‘s the music of the Oruro region.

4. When Is Carnival?
(a) It changes from year to year.
(b) One month after Easter.
(c) Two weeks after Christmas.
(d) The fourth month of the year.

5. Where do most of the people work in Oruro?
(a) In a factory where musical instruments are made.
(b) In the jungle in Bolivia.
(0) In the mines in Oruro.
(d) In a store where costumes are made.

 

141

6. What metal was extracted in large quantities from an area

10.

11.

12.

near Oruro?
(a) Gold

(b) Silver . ~-
(C) Copper
(d) Zinc

Who was Nusca?

(a) A figure brought by the conquistadors.
(b) A goddess from Greek mythology.

(c) A princess who saved her people.

((1) The saint that protects the Oruro region.

Why did Huari want revenge on the Urus?
(a) Because the Urus changed their lifestyle.
(b) Because the Urus worshipped serpents.
(0) Because the Urus declared war on Huari.
(d) Because Huari did not like animals.

Why were the animals turned into stone and sand?
(a) So that they could not attack the Urus.

(b) Because they did not obey Huari.

(0) Because they turned into gods for the Incas.
(d) Because Huari wanted revenge on the Urus.

What country took advantage of the natural resources of the
Oruro area?

(a) England.

(b) Germany.

(c) Portugal.

(d) Spain.

Where is Oruro?

(a) Perhaps in a flat region.

(b) Perhaps in the jungle.

(c) Perhaps In the desert.

(d) Perhaps in a mountainous region.

What can we say about today's miners of the Oruro region?
(a) They seem to be religious people.

(b) They make their own costumes for the festival.

(c) They still worship Greek gods.

(d) They wear bright colors only for this festival.

1 42
COMPREHENSION ASSESSMENTUNFAMILIAR TOPIC

(B) .............................................

The questions below are about La Diablada . Read them carefully
and choose one item 'a,‘ 'b,‘ 'c,‘ or 'd' that answers each question
correctly

1. What is the origin of La Diablada?
(a) It's a mixture of myth and religion.
(b) It's the history of a group of native Indians.
(0) It's completely unknown.
(d) It‘s the music of the Oruro region.

2. Where do people celebrate La Diablada?
(a) In a Caribbean country.
(b) In a country in South America.
(0) In a region in Mexico.
(d) On some Spanish islands.

3. What can we say about today's miners of the Oruro region?
(a) They seem to be religious people.
(b) They make their own costumes for the festival.
(c) They still worship Greek gods.
(d) They wear bright colors only for this festival.

4. When is Carnival?
(a) It changes from year to year.
(b) One month after Easter.
(c) Two weeks after Christmas.
(d) The fourth month of the year.

5. Where do most of the people work in Oruro?
(a) In a factory where musical instruments are made.
(b) In the jungle in Bolivia.
(c) In the mines In Oruro.
(d) In a store where costumes are made.

143

6. What is La Diablada?

10.

11.

12.

(a) A group of people

(b) A celebration

(c) A popular dance . ~-
(d) A type of music

Who was Nusca?

(a) A figure brought by the conquistadors.
(b) A goddess from Greek mythology.

(c) A princess who saved her people.

(d) The saint that protects the Oruro region.

Why did Huari want revenge on the Urus?
(a) Because the Urus changed their lifestyle.
(b) Because the Urus worshipped serpents.
(c) Because the Uan declared war on Huari.
(d) Because Huari did not like animals.

Why were the animals turned into stone and sand?
(a) So that they could not attack the Ums.

(b) Because they did not obey Huari.

(c) Because they turned Into gods for the Incas.
(d) Because Huari wanted revenge on the Urus.

What metal was extracted in large quantities from an area
near Oruro?

(a) Gold

(b) Silver

(C) 00999?

(d) Zinc

Where is Oruro?

(a) Perhaps in a flat region.

(b) Perhaps in the jungle.

(c) Perhaps in the desert.

((1) Perhaps in a mountainous region.

What country took advantage of the natural resources of the
Oruro area?

(a) England.

(b) Germany.

(c) Portugal.

(d) Spain.

APPENDIX 8
ADDITIONAL HISTOGRAMS FOR RECALLS AND COMPREHENSION

APPENDIX B
HIST OGRAMS FOR RECALLS AND COMPREHENSION

.—-——_—_————-————————————-——————-——_—————————_————-_

 

Midpoint Count
2 1 *
3 0
4 2 * *
5 5 .....
6 10 ..........
7 13 .............
8 9 tktflfl'fl‘k'kw
9 6 *R'KW‘R‘R
10 12 ............
11 3 " "‘ *
12 1 "
FIGURE 13

' COMPREHENSION SCORES FOR UNFAMILIAR TEXT

 

N= 62 bckgrnd: 1

 

Midpoint Count

 

 

2 0

3 0

4 1 "

5 4 WWW!

6 4 Wk‘kt

7 6 WWW!!!

8 7 Witt!!!

9 18 RXW‘RWRWWKXIXW'RWWWX

10 9 wxxxwwwaw

11 12 *k‘kt‘kfl‘kflflflkfl
_ 12 1 *

FIGURE14

COMPREHENSION SCORES FOR FAMILJAR TEXT
1 44

145

 

N=63 vcbdiff: 0
Midpoint Count

 

2 1 '
3 0
4 1 *
5 6 Xkflltk
6 7 .4..*..
7 1O WWIKRW‘RWKR
8 7 xwxxwrx
9 14 wwwxxxaawxxaas
1O 9 wasnxsanw
11 8 kttflflflkt
12 O

HGURE15

COMPREHENSION SCORES FOR HARD VERSION

 

N= 62 vcbdiff: 1
Midpoint Count

 

 

2 o

3 o

4 2 **

5 3 WWW

6 7 WWWR'A’XW

7 9 ‘k‘kfll’kflkkfl

8 9 Wkkkflktflfl

9 10 Wflfl‘kfl'ktflflfl

10 12 ttfl‘kkt'kflikx‘k

11 7 *‘k‘kflk‘k'k

12 2 **
HGURE16

COMPREHENSION SCORES FOR EASY VERSION

146

 

N= 62 CtrI-ld bcqund: 0
Midpoint Count

 

 

0 1O *tflt‘kl’flfllfl
1 26 *Iflflfi'tfllflfl‘ktflkfll*WWRWXRWRR
2 12 HH...HHH
3 9 *xwrwwrar
4 2 "*
5 2 ""
6 0
7 1 "
FIGURE17

CENTRAL IDEAS OF UNFAMILIAR TEXT

 

N= 62 CtrI-ld bckgmd= 1

 

 

 

Midpoint Count
0 15 wasxwaxxwwxxwwx
1 15 ...............
2 17 xxxawasxwxxwwwsxx
3 10 ..x.......
4 4 ....
5 1 *
6 0
7 1 *
HGURE18

CENTRAL IDEAS OF FAMILIAR TEXT

147

 

N=62 Spring-Id bckgrnd: 0 _,
Midpoint Count

 

O 11 Iflflklkfiflttl
1 15 *Rflflxkfl‘kklt‘kflflk
2 23 *WRWIKWWWIWXWIWWWXIXIRW
3 10 WIIRKKWWRA
4 2 *"
5 1 *
6 O
FIGURE19

SUPPORTING IDEAS OF UNFAMILIAR TEXT

 

N=62 Spring—Id bcgnd=1
Midpoint Count

II

**K**R*fl*
****************X
W****XWW*RWWWW*XRW*‘R
*WW‘R‘R

*fiflfl‘kflkw

‘R

0:01.503on
”-3
Ammo-10”

 

FIGURE 20
SUPPORTING IDEAS OF FAMILIAR TEXT

148

 

N=62 CtrI-ld vcbdiff: 0
Midpoint Count

 

O 15 *Kfltkflfltflflflflfflfl
1 19 IRWWRIWI*IW****W***
2 20 *****W*****R****W**fl
3 8 IXRWWWWW
4 1 *
5 O
6 O
7 O
FIGURE21

CENTRAL IDEAS OF HARD VOCABULARY

 

N= 62 CtrI-Id VdeIII= 1

 

 

 

Midpoint Count
0 10 usuaxaarwx
1 22 waxwxwwwwwwxwwxwwwsxrw
2 9 waxwwwnxw
3 11axwwsw~kwwxx
4 5 sxwww
5 3 «an
6 O
7 1*
FIGURE22

CENTRAL IDEAS OF EASY VOCABULARY

149

 

N= 62 Sprtng-Id vcbdiff: 0

 

 

 

Midpoint Count ‘“
I
0 11 nwtnwwtwwww
1 13 wwwwxwrwwwwxw
2 23 wwwxxwwwwwxxxwwwtwrwwtw
3 9 xxwwwxxww
4 2 **
5 4 xxx:
6 1 *
FIGURE23

SUPPORTING IDEAS OF HARD VOCABULARY

 

N= 62 Sprtng-ld vcbdiff: 1

 

 

 

Midpoint Count
0 2 «w
1 11 ltflttktttkt
2 17 *wwwxwwnrwwwwwwww
3 21 wanwwwwwxxxwwwwwwxwww
4 5 Wit!!!
5 5 3'3?wa
6 0
FIGURE24

SUPPORTING IDEAS OF EASY VOCABULARY

150

 

N=62 Explt-Qs bckgrnd: 0

 

 

Midpoint Count
0 0
1 3 kirk
2 15 wwwxxnxxwwwwwxw
3 23 waxwuwxxwxxnwtnwxkawwx
4 21wkuunuwxxwuwwnnwwt~xwu

 

FIGURE 25
EXPLICIT QUESTIONS OF UNFAMILIAR TEXT

 

N= 62 Eprt-Qs bckgrnd=1

 

Midpoint Count

2 I X

6 *kl’k‘kfl
12 **K*****W***

16 flflfltlkflflkflkfik‘k‘kk

AQN-‘O

26 I********II*K******‘R'kfiwtl‘k

 

FIGURE 26
EXPLICIT QUESTIONS OF FAMILIAR TEXT

 

N= 62 lmpl-Qs bckgrnd: 0
Midpoint Count

3!!!

9 ********‘R
22 *Xxflfllfik'kfififlkflkflk'kflkWt

19 *Rktkwflflkflflflflkfiflflflfl

AWN-*0

9 flflflkflflkflfl

 

FIGURE 27
IMPLICIT QUESTIONS OF UNFAMILIAR TEXT

151

 

N= 62 impI-Os bckgmd=1

 

 

 

Midpoint Count
0 2 **
1 16 wwwwtwwtwwwxxwww
2 21xwwrwwwrwawwxwwwwwxwk
3 21*fl‘ktflktfl‘kflkkkt‘kflflflkflw
4 2 **
FIGURE28

IMPLICIT QUESTIONS OF FAMILIAR TEXT

 

N= 62 Scrpl-Qs bckgrnd: 0
Midpoint Count

1*
12 *fl'flkflkfl‘k‘kfi‘kfl

21*WflKflWWRWIWWKKK‘k‘kXXKK

21 *fltkflflflkflk*flflfiflfl***flk

hQN-‘O

7 Ifl‘ktfltl

 

FIGURE 29
SCRIPTICALLY IMPLICIT QUESTIONS OF UNFAMILIAR TEXT

 

N= 62 ScrpI-Qs bckgrnd: 1
Midpoint Count

 

0 O
1 O
2 1 "‘
3 15 *W'kk‘k‘kfl‘kWWWR‘kflk
4 46 XWXW’tWWwflkkflflflfiwflkflkflflfi'
****fl*W**X*fl****X**X*WX
FIGURE30

SCRIPTICALLY IMPLICIT QUESTIONS OF FAMILIAR TEXT

152

 

N=62 Eprt-Qs vcbditt: 0
Midpoint Count

2 **

6 ****W*
15 *Wkflkkfll’flfllflkl'W

19 fl*****X****!**RW*FW

hOJN-‘O

21*Rfifltflflktflkflk‘kflflkflflfifl

 

FIGURE 31
EXPUCIT QUESTIONS OF HARD VOCABULARY

 

N: 62 Explt-Qs vcbdiff: 1
Midgoint Count

0
3***

12 *****X*K**X*

20 *WXXflWKflR‘R‘KRRflWWWtXW

#(aJN-SO

26 *‘k‘k********WflflWRkfiwflkI’KWK'fl

 

FIGURE 32
EXPLICIT QUESTIONS OF EASY VOCABULARY

 

N= 62 lmpI-Qs vcbdiff: 0
Midpoint Count

1*
16 *Kflflflfl‘kflkflfiflwflfi'k

21******W‘kflfl******fl*flflk

20 RWI’XWXXXWII‘fiW‘R‘Rwflfiflfl

bum—*0

5 *‘k‘k‘k!

 

FIGURE 33
IMPLICIT QUESTIONS OF HARD VOCABULARY

153

 

N= 62 Impl-Qs vcbdiff: 1
Midpoint Count

4 ”I!“

9 *Kfll’l’tlfl‘k

*flk****fl***********flk*

20 *flflflkfl!*l***k*****fl*

«AON-‘O
N
N

6 *Kflflfll

 

FIGURE 34
IMPLICIT QUESTIONS OF EASY VOCABULARY

 

N= 62 Scrpl-Qs vcbdiff: O
Midpoint Count

1*
6 KKK!!!

10 IXRIRKRWXfl

17 *fikkflkttlflkfl‘kfltflfl

«bQN—‘O

29 Itfitflflflflkkflflfl‘kflfl‘kflflltflfl

Ikkfltfl

 

FIGURE 35
SCRIPTICALLY IMPLICIT QUESTIONS OF HARD VOCABULARY

 

N= 62 ScrpI-Qs vcbditt: 1
Midpoint Count

0

6 *X'kflflk
12 *flflkl’flk‘k‘kflkk

19 Rkflflkkflfifl‘ktkflkfltflflk

AQN-‘O

25 *R*****fl**fl*‘k*flk**fl******

 

FIGURE 36
SCRIPTICALLY IMPLICIT QUESTIONS OF EASY VOCABULARY

APPENDIX C
ADDITIONAL STATISTICAL ANALYSIS OF THE DATA

APPENDIX C

ADDITIONAL STATISTICAL ANALYSIS OF THE DATA

 

 

 

 

 

 

TABLE 30
MULTIVARIATE ANALYSIS OF COVARIANCE FOR AGE

MA—NCGVA for age _____ s; _—1_—_ n'T: 1.5 —————— 6" =—--5-33 —————
CRTT-E—R—IC—DN_-————TE—S-T_STATI_STI_C """" E "“"6F"’""T=""
Wilk's 0.98813 0.262 (5, 109) 0.933
Lawley-Hotelling 0.01202 0.262 (5. 109) 0.933
Pillai's 0.01187 0.262 (5, 109) 0.933
Roy's 0.01202

TABLE 31

MULTIVARIATE ANALYSIS OF COVARIANCE FOR GENDER

9171360; 73? 536a}? ”””””” E: "T ““““ .3 21'75 """" H 353""
Efifi‘e‘n‘iafi’"""iE'é’T‘s’fifi’é'fic ‘ F “B?- """"""" p "‘—
Wilk's 0.97260 0.614 (5, 109) 0.689
Lawley-Hotelling 0.02818 0.614 (5. 109) 0.689
Pillai's 0.02740 0.614 (5, 109) 0.689
Roy's 0.02818

TABLE 32

MULTIVARIATE ANALYSIS OF COVARIANCE FOR YEARS OF LANGUAGE
STUDY
MIN—669376?93513617555355; ‘s‘t‘JaQ ““““““““““““““““
s: 1 m= 1.5 n: 53.5

CRITERION TEST STATISTIC F OF P
Wilk's 0.92081 1.875 (5, 109) 0.105
Lawley-Hotelling 0.08600 1.875 (5. 109) 0.105
Pillai's 0.07919 1.875 (5, 109) 0.105
Roy's 0.08600

_————————-——_—————————.——————————.——————.—_—

————_—-——————-_————_

155

TABLE 33
MULTIVARIATE ANALYSIS OF COVARIANCE FOR SPANISH-SPEAKING
FRIENDS/RELATIVES , ~-
M-A-ITC—OTI-A-T‘Sr Spanish- -sp68_IITrTg——tTI—endslrelatives ———————————————
s: 1 In: 1. 5 n= 53.5

 

 

CRITERION TEST STATISTIC F DF P
wmcs 0.96249 0.850 (5, 109) 0.518
Lawley-Hotelling 0.03897 0.850 (5. 109) 0.518
PiIIaI‘s 0.03751 0.850 (5, 109) 0.518
Roy's 0.03897

TABLE 34

MULTIVARIATE ANALYSIS OF COVARIANCE FOR STUDENTS' VISITS TO
A SPANISH SPEAKING COUNTRIES

MANCOVA for visit a Spanish- speaking country
s: 1 m: 1.5 n: 53.5

 

 

 

 

CRITERION TEST STATISTIC F DF P
Wilk's 0.97226 0.622 (5, 109) 0.683
Lawley-Hotelling 0.02853 0.622 (5. 109) 0.683
Pillai's 0.02774 0.622 (5, 116) 0.683
Roy's 0.02853

TABLE 35

MULTIVARIATE ANALYSIS OF COVARIANCE FOR LENGTH OF STAY IN
SPANISH SPEAKING COUNTRIES

 

—_-———-—_-——_——_———__——_—————_————————_-

MANCOVA for stay in Spanish-speaking country

CRITERION TEST STATISTIC F DF P
Wilk's 0.95379 1.056 (5, 109) 0.389
Lawley-Hoteiling 0.04845 1.056 (5. 109) 0.389
Pillai's 0.04621 1.056 (5, 116) 0.389

Roy's 0.04845

——-_—_—*—_———_—————-———-_——————~——_——_——————————_———_—fl————_——

156

TABLE 36
MULTIVARIATE ANALYSIS OF COVARIANCE FOR CONTACT WITH THE
LANGUAGE OUTSIDE THE CLASSROOM . 1"

 

 

 

 

 

 

 

 

 

 

MANCOVA for language contact & 1 m: 1.5 n.—. 53.5
CRTT—E-R-ION-——-_-TE—ST—STA—TI-STTC ————————— E ______ ISF— ——————— R —
Wilk's 0.96826 0.715 (5, 109) 0.614
Lawley-Hotelling 0.03279 0.715 (5. 109) 0.614
Pillai's 0.03174 0.715 (5, 109) 0.614
Roy's 0.03279

TABLE 37

MULTIVARIATE ANALYSIS OF COVARIANCE FOR BACKGROUND

KNOWLEDGE
MA-INTCOQ-A—TOI—TOpTO—Tafifiarity _s;"1—"Ffi:_1_§"—H=- “535"—
CRIT_ERION_————-TE—ST_S—TATISTI_C _________ E —————— 5E _______ R _
Wilk's 0.40381 32.186 (5, 109) 0.000
Lawley-Hotelling 0.47640 32.186 (5. 109) 0.000
Pillai's 0.59619 32.186 (5, 109) 0.000
Roy's 1.47640

TABLE 38

MULTIVARIATE ANALYSIS OF COVARIANCE FOR VOCABULARY

MA-N—COQA-TOFNEB-OUTa—ry ————————————— s; 1 m= 1‘5"];"5-35—
BREE—RON —————— TEST STATISTTO _____ F “nu—"NOT: ________ R —
Wilk's 0.91606 1.997 (5, 109) 0.085
Lawley-Hotelling 0.09163 1.997 (5. 109) 0.085
Pillai's 0.08394 1.997 (5, 116) 0.085

Roy's 0.09163

_‘—_——_————-——_——————_————.——_-————_—————_——————_———_—*———_——

157

TABLE 39
MULTIVARIATE ANALYSIS OF COVARIANCE FOR INTERACTION BETWEEN
BACKGROUND AND VOCABULARY . ...

 

RAN—OOI—I'IYSr—YSfiE—IamiIIamy—

and vocabulary interaction s: 1 m: 1.5 n:
53.5
CRITERION TEST STATISTIC F DF P
Wilk's 0.96444 0.804 (5, 116) 0.549
Lawley-Hotelling 0.03687 0.804 (5. 116) 0.549
Pillai's 0.03556 0.804 (5, 116) 0.549

Roy's 0.03687

_——~———__-_-—_——_———_————_——————_——_——--—~———_——_——-—————_

IJST OF REFERENCES

LIST OF REFERENCES

Alba, J. W., & Hasher, L. (1983). Is memory schematic? Psychological
Bulletin, 93, 203-231.

Alderson, J. C., & Urquhart, A. H. (1984). Reading in a foreign
language. Essex, England: Longman Group Limited.

Alverrnann, D. E., Smith, L. C., & Readence, J. E. (1985). Prior
knowledge activation and the comprehension of compatible text.
Reading Research Ouarterl y, 20, 420-436.

Anderson, R. C., & Freebody, P. (1981). Vocabulary knowledge. In J.
Guthrie (Eds), Comprehension and teaching: Research reviews.
Newark, Delaware: lntemational Reading Association.

Anderson, R. C., & Pearson, P. D. (1989). A schema-theoretic view of
basic processes. In P. L. Carrell, J. Devine, & D. E. Eskey (Eds),
interact! ve approaches to second language reading. New York:
Cambridge University Press.

Anderson, R. C., Pichert, J. W., & Shirey, LL (1983) Effect of the
reader's schema at different points in time. Joumal of Educational
Psychology, 75, 271-279.

Anderson, R. C., Reynolds, R. E., Schallert, D. L., & Goetz, E. T. (1977).
Frameworks tor comprehending discourse. American Educational
Research Journal, 14, 367-381. '

Aron, H. (1983). The influence of background knowledge on memory
for reading passages by native and nonnative readers. TESOL
Quarterly, 20, 136-140.

Asher, S. R., Hymel, 8., & Wigfield, A. (1978). Influence of topic
interest on children reading comprehension. Journal of Reading
Behavior, 10, 35-47.

158

159

Balota, D. A., Ferraro, F. R, 8: Connor, L. T. (1991 ). On the early
influence of meaning in word recognition: A review of the literature.
In P. J. Schwanenflugel, The psychology of word meanings. Hillsdale,
NJ: Lawrence Erlbaum Associate, Publishers.

Barnett, Marva. (1989). More than meets the eye. New Jersey:
Prentice-Hall.

Bamitz, J. G. (1986). Toward understanding the effects of cross-
cultural schemata and discourse structure on second language
reading comprehension. Journal of Reading Behavior, 28, 95-116.

Beck, I. L., & Juel, C. (1992). The role of decoding in learning to read.
In Samuels, S. J., & A. E. Farstrup (Eds), What research has to say
about reading instruction. Newark, DE: International Reading
Association.

Bellonl, L. F., & Jongsma, E. A. (1978). The effect of interest on
reading comprehension of low-achieving students. Joumal of
Reading. 22, 106-109.

Bhatia, V. K. (1984). Syntactic discontinuity in legislative writing
for academic legal purposes. In A. K. Pugh, & J. M. Ulijn (Eds),
Reading for professional purposes: Studies and practices in native
and foreign language. London: Heinemann Educational Books.

Blau, E. K. (1982). The effect of syntax on readability for ESL
students in Puerto Rico. TESOL Ouarterty, 16, 517-528.

Brewer, W. F. & Treyens, J. C. (1981). Role of schemata in memory for
places. Cognitive Psychology, 13, 207-230.

Brown, A. L., & Smiley, S. S. (1977). Rating the importance of
structural units of prose passages: A problem of metacognitlve
development. Child Development, 48, 1-8.

Caimey, T. H. (1990). Teaching reading comprehension. meaning
markers at work. Bristol, PA: Open University Press.

160

Carey, R., Harste, J., & Smith, S. (1981). Contextual constrains and
discourse processes: A replication study. Reading Research
Quarterly, 16, 201-212. . ~-

Carrell, P. L. (19838). Three components of background knowledge in
reading comprehension. Language Leaming, 33, 183-208.

Carrell, P. L. (19830). Background knowledge in second language
comprehension. Language Learning and Communication 2, 25-34.

Carrell, P. L. (1984a). The effect of rethorical organization on ESL
readers. TESOL Quarteriy, 18, 441-469.

Carrell, P. L. (19840). Evidence of a formal schema in second
language comprehension. Language Leamlng, 34, 87-112.

Carrell, P. L. (1987). Content and formal schemata in ESL reading.
TESOL Quarterly 21, 461-481.

Carrell, P. L 8. Eisterhold, J. C. (1983). Schema theory and ESL
reading pedagogy. TESOL Quarterfy 18, 553-573.

Carrell, P. L., & Eisterhold, J. C. (1989). Schema theory and ESL
reading pedagogy. In P. L Carrell, J. Devine, 8 D. E. Eskey (Eds),
lnteracfi ve approaches to second language reading. New York:
Cambridge University Press.

Carver, R. P. (1977). Toward a theory of reading comprehension and
rauding. Reading Research Quarterly, 13, 8-63.

Carver, R. P. (1978). Sense and nonsense about generalizing to a
language pOpulatlon. Journal of Reading Behavior, 10, 25-33.

Carver, R. P. (1990). Reading rate. A review of research and theory.
San Diego: Academic Press, Inc.

Carver, R. P. (1992). Effect of prediction activities, prior knowledge,
and text type upon amount comprehended: Using rauding theory to
critique schema theory research. Reading Research Quarterly, 27,
165-174.

161

Chiarello, C. (1991). Interpretation of word meanings by the cerebral
hemisphere: One is not enough. 1991. In P. J. Schwanenflugel, The
psychology of word meanings Hillsdale, NJ: Erlbaum Associate, --
Publishers.

Connor, U. (1987). The eclectic synergy of methods of reading
research. In J. Devine, P. L. Carrell, and D. E. Eskey (Eds), Research in
reading in English as a second language. Washington, D. C.: TESOL.

Cunningham,J. W., Cunningham, P. M., & Arthur, S. V. (1981 ). Middle
and secondary school reading. New York: Longman.

Cunningham, J. W., & Moore, D. W. (1993). The contribution of
understanding academic vocabulary to answering comprehension
questions. Journal of Reading Behavior, 25 (2), 171-180.

Cunningham, J. W., & Wall, L. K. (1994). Teaching good readers to
comprehend better. Joumai of Reading, 37 (6), 480-486.

Danan, M. (1992). Reversed subtitling and dual coding theory: New
directlons.for foreign language instruction. Language Leamlng, 42
(4), 497-527.

Davis, J. N. (1989). Facilitating effects of marginal glosses on
foreign language reading. The Modern Language Joumai, 73, 41-48.

Denner, P. R. (1992). Effect of prereading activities on junior high
students‘ recall. Joumai of Educational Research, 86, 11-19.

Devine, J. (1987). General language competence and adult second
language reading. In J. Devine, P. L. Carrell, and D. E. Eskey (Eds),
Research in reading in English as a second language. Washington, D.
C.:TESOL '

Devine, J., Carrell, P. L., & Eskey, D. E. (1987). Research in reading in
English as a second language. Washington, D. C.: TESOL.

Dole, J. A., Valencia, S. W., Greer, E. A., 8. Wardrop, J. L. (1991).
Effects of two types of prereading instruction on the comprehension
of narrative and expository text. Reading Research Quarterly, 26,
142-158.

162

Dubin, F., Eskey, D. E., & Grabe, W. (1986). Teaching second language
reading for academic purposes. Reading, Massachusetts: Addison--
Wesley Publishing Company.

Eskey, D. E. (1989). Holding in the bottom: An interactive approach to
the language problems of second language readers. In P. L. Carrell, J.
Devine, & D. E. Eskey (Eds), interactive approaches to second
language reading. New York: Cambridge University Press.

Gambrell, L. B., & Brooks Jawitz, P. (1993). Mental imagery, text
illustrations, and children's story comprehension and recall. Reading
Research Quarterly, 28, 265- 276.

Garner, R. (1992). Metacognition and self-monitoring strategies. In
Samuels, S. J. & A. E. Farstrup, What Research has to say about
reading instruction. Newark, DE: lntemational Reading Association.

Gentine, L. M., Kamil, M. L, & Blanchard, J. S. (1983). Reading
research revisited. Columbus, Ohio: Charles E. Merrill Publishing
Company. .

Goodman, K. (1968). Reading: a psycholinguistic guessing game.-
Journal Of the Reading Specialist, 6, 126-135.

Goodman, K. (1981). Miscue analysis and future research directions.
In S. Hudelson (Eds), Learning to reading in different languages.
Linguistic and literacy series. Washington, DC: Center for Applied
Linguistics.

Goodman K. (1989). The Reading process. In P. L Carrell, J. Devine, &
D. E. Eskey (Eds), interactive approaches to second language reading.
New York: Cambridge University Press.

Goodman, K. (1992). Whole language research: Foundations and
development. In S. J. Samuels, & A. E. Farstrup (Eds), What research
has to say about reading instruction. Newark, DE: International
Reading Association.

163

Grabe, W. (1989). Reassessing the term interactive. In P. L Carrell, J.
Devine, & D. E. Eskey. (Eds), interactive approaches to second
language reading. New York: Cambridge University Press. , ~-

Guarino, R., & Perkins, K. (1986). Awareness of form class as a
factor In ESL reading comprehension. Language Leamlng, 36, 77-82.

Guilford, J. P. & Fruchter, B (1978). Fundamental statistics in
psychology and education . New York: McGraw-Hill, Inc.

Hammadou, J. (1991). Interrelationships among prior knowledge,
Inference, and language proficiency In foreign language reading. The
Modern Language Joumai, 75, 27-38.

Harrison, C. (1992). The reading process and learning to read: What a
teacher using a 'Real Book' approach needs to know. In Harrison, 0., &
M. Coles. (Eds), The reading for real handbook. London: Routledge.

Harrison, 0., & Coles, M. (1992). The reading for real handbook.
London: Routledge.

Hayes, J. R. (1981 ). The complete problem solver. Hillsdale, NJ:
Lawrence Erlbaum.

Henk, W. A., Stahl, N. A., & Melnlck, S. A. (1993). The Influence of
readers‘ prior knowledge and level of involvement on ambiguous text

interpretation: An extension study. Reading Research and instruction,
32, 1-12.

Hiebert, E. H., & Calfee, R. C. (1992). Assessing literacy: From
standardized tests to portfolios and performances. In S. J. Samuels,
& A. E. Farstrup (Eds), What research has to say about reading
instruction. Newark, DE: lntemational Reading Association.

Hogben, D., & Lawson, M. J. (1994). Keyword and multiple elaboration
strategies for vocabulary acquisition in foreign language learning.
Contemporary Educational Psychology, 19, 367-376.

Huckin, T., Haynes, M., & Coady, J. (1993). Second language reading
and vocabulary leamlng. New Jersey: Ablex Publishing Corporation.

164

Iran-Nejad, A. (1987b). The schema: A Iong-terrn memory structure

or a transient functional pattern. In R. J. Tierney, P. L. Anders, & J. N.
Mitchell (Eds), Understanding reader‘s understanding (pp. 109-127)
Hillsdale, NJ: Erlbaum.

lnNin, J. w. (1991). Teaching reading comprehension processes. New
Jersey: Prentice Hall.

James, M. O. (1987). ESL reading pedagogy: Implications of schema-
theoretical research. In J. Devine, P. L. Carrell, & D. E. Eskey (Eds),
Research in reading in english as a second language. Washington, D.
C.:TESOL

Johnson, Patricia. (1982). Effects of reading comprehension on
building background knowledge. TESOL Quarterly, 16, 503-516

Johnson, D. D. & Pearson, P. D. (1984). Teaching reading vocabulary.
New York: Holt, Rinehart & WInston.

Johnston, P. H. (1983). Reading comprehension assessment: A
cognitive basis. Newark, DE: lntemational Reading Association.

Just, M. A. & Carpenter, PA. (1987). The psychology of reading and
reading comprehension. Massachusetts: Allyn and Bacon, Inc.

Kim, S. (1995). Types and sources of problems in L2 reading: A
qualitative analysis of the recall protocols by Korean high school
EFL students. Foreign Language Annals, 28, 49-70.

Klein, M. L. (1988). Teaching reading comprehension and vocabulary: A
guide for teachers. New Jersey: Prentice Hall.

Koda, K. (1987). Cognitive strategy transfer in second language
reading. In J. Devine, P. L Carrell, & D. E. Eskey (Eds), Research in
reading in English as a second language. Washington, D. C.: TESOL.

Koda, K. (1989). The effects of transferred vocabulary knowledge on
the development of L2 reading proficiency. Foreign Language Annals
22, 529-540.

165

Koda, K. (1992). The effect of lower-level processing skills on FL
reading performance: Implications for instruction. The Modern ~-
Language Joumai, 76, 502-512.

Krashen, S. D. (1989). We acquire vocabulary and spelling by reading.
The Modern Language Joumai, 73, 440-464.

Krashen, S. D. (1993). The power of reading. insights from the
research. Colorado: Libraries Unlimited, Inc.

Lee, J. (1986). Background knowledge 8 L2 reading. The Modern
Language Joumai, 70, 350-354.

Larsen-Freeman, D., & Long, M. H. (1991). An introduction to second
language acquisition research. London: Longman.

Laufer, B. (1990). Ease and difficulty in vocabulary learning: Some
teaching implications. Foreign Language Annals, 23, 147-155.

Lee, J. (1987). FL Learners’ ability to recall and rate the important
ideas of an expository text. In B. VanPatten, T. R. Dvorak, & J. F. Lee
(Eds), Foreign language leamlng. A research perspective.
Cambridge: Newbury House Publishers.

Lee, J. F., & Riley, G. L. (1990). The effect of prereading,
rhetoricalIy-oriented frameworks on the recall of two structurally

different expository texts. Studies in second language acquisition,
12, 25-41.

Littlefair, A. (1992). Reading and writing across the curriculum. In
Harrison, 0., & M. Coles. (Eds), The reading for real handbook. London:
Roufledge.

Long, S. A., Winograd, P. N., & Bridge, C. A. (1989). The effect of
reader and text characteristics on reports of imagery during and
after reading. Reading Research Quarterly, 24, 353-372.

Malt. B. C. (1991 ). Word meaning and word Use. In P. J.
Schwanenflugel (Eds), The psychology of word meanings. Hillsdale,
NJ: Erlbaum Associate, Publishers.

166

Mayer, R. E., & Anderson, R. B. (1991). Animations need narrations: An
experimental test of a dual coding hypothesis. Joumai of Educational
Psychology, 83, 484-490.

Mayer, R. E., 8. Anderson, R. B. (1992). The instructive animation:
helping students build connections between words and pictures in
multimedia learning. Joumai of Educational Psychology, 84, 444-
452.

McKeown, M. G. (1991). Learning word meanings from definitions:
problems and potential. In P. J. Schwanenflugel (Eds), The
psychology of word meanings. Hillsdale, NJ: Erlbaum Associate,
Publishers.

McKeown, M. G., Sinatra, G. M., & Loxterrnan, J. A. (1992). The
contribution of prior knowledge and coherent Text to comprehension.
Reading Research Quarterly, 27, 79-93.

Medo, M. A., & Ryder, R. J. (1993). The effects of vocabulary
instruction on reader's ability to make causal connections. Reading
Research and lnsthCtlon, 33 (2) 119-134.

Merchant, G. (1992). Supporting readers for whom English is a second
language. In C. Harrison, & M. Coles. (Eds), The reading for real
handbook. London: Routledge.

Moore, D. W., & Moore, S. A. (1986). Possible sentences. In E. K.
Dishner, T. W. Bean, J. E. Readance, & D. W. Moore (Eds), Reading in
the content areas: improving classroom instnlction. Dubuque, IA:
Kendalll Hunt.

Murphy, G. L. (1991). Meaning and concepts. In P. J. Schwanenflugel
(Eds), The psychology of word meanings. Hillsdale, NJ: Erlbaum
Associate, Publishers.

Nelson G. L. (1987). Culture's role in reading comprehension: A
schema theoretical approach. Joumai of Reading, 30 (5), 425-429.

167

Nurse, J. R, & Hough, R. A. (1992). Reading and the ESL student. In S.
J. Samuels, & A. E. Farstrup (Eds), What research has to say about
reading instruction. Newark, DE: lntemational Reading Association.

Oakhill, J., & Gamham, A. (1988). Becoming a skilled reader. New
York: Basil Blackwell Ltd.

Omanson, R. C. (1982). An analysis of narratives: Identifying central,
supportive, and distracting content. Discourse Processes, 5, 195-
224.

Otto, W. (1983). A commentary on LaBerge and Samuels'
retrospective analys’s of the theory of automaticity In reading. In L.
Gentile, M. Kamll & J. Blanchard (Eds), Reading research revisited.
Columbus, Ohio: Charles E. Merrill Publishing Company.

Otto, W. (1992). The role of research In reading instruction. In S. J.
Samuels and A. E. Farstrup (Eds), What research has to say about
reading instruction. Newark, DE: lntemational Reading Association.

Ousbey, J. (1992). Reading and the Imagination. In Harrison, 0., & M.
Coles (Eds), The reading for real handbook. London: Routledge.

Paivio, A. (1986). Mental representations: A dual coding approach.
New York: Oxford University Press.

Parry, K, J. (1987). Reading in a second culture. In J. Devine, P. L.
Carrell, & D. E. Eskey (Eds), Research in reading in English as a
second language. Washington, D. C.: TESOL.

Pearson, P. D., Roehler, L. R., Dole, J. A., & Duffy, G. G. (1992). In S. J.
Samuels, & A. E. Farstrup (Eds), What research has to say about
reading instruction. Newark, DE: lntemational Reading Association.

Peeck, P. D., van den Bosch, A. B., & Kreupeling, W. J. (1982). Effect of
mobilizing prior knowledge on learning from text. Joumai of
Educational Psychology, 74, 771-777.

Perkins, K. (1987). The Relationship between nonverbal schematic
concept formation and story comprehension. In J. Devine. P. L.

168

Carrell, & D. E. Eskey (Eds), Research in reading in English as a
second language. Washington, D. C.: TESOL.

Pichert, J. W., & Anderson, R. C. (1977). Taking different
perspectives on a story. Joumai of Educational Psychology, 69, 309
315.

Prefiey, M., Levin, J. R., & Delaney, H. D. (1982). The mnemonic
keyword method. Review of Educational Research, 52, 61-91.

Pritchard, R. (1990). The effects of cultural schemata on reading
processing strategies. Reading Research Quarterly, 25, 273-295.

Rumelhart, D. (1977). Toward an interactive model of reading. In S.
Domic (Eds), Attention and performance. New Jersey: Erlbaum.

Rumelhart, D. (1980b). Schernata: The building block of cognition. In
R. J. Spiro, & W. F. Brewer (Eds), Theoretical issues in reading
comprehension. Hillsdale: Erlbaum.

Rumelhart, D. E., 8 Ortony, A. (1977). The representation of
knowledge in memory. In R. C. Anderson, R. J. Spiro, & W. E. Montague
(Eds), Schooling and the acquisition of knowledge. Hillsdale:
Erlbaum.

Sadoski, M. (1983). An exploratory study of the relationships
between imagery and the comprehension and recall of a story.
Reading Research Quarterly, 19, 110-123.

Sadoski, M., Goetz, E. T., Olivarez, A. , Lee, S., & Roberts, N. M. (1990).
Imagination in story reading: The role of imagery, verbal recall,
story analysis, and processing levels. Joumai of Reading Behavior,
22, 55-70.

Sadoski, M., Goetz, E. T., & Fritz, J. B. (1993). Impact of concreteness
on comprehensibllity, interest, and memory for text: Implications
for dual coding theory and text design. Joumai of Educational
Psychology, 85, 291-304.

Sadoski, M., Goetz, E. T., & Fritz, J. B. (1993). A causal model of
sentence recall: Effects of familiarity, concreteness,

169

comprehensibllity, and interestingness. Joumai of Reading Behavior,
25, 5-16.

Sadoski, M., Paivio, A., 8 Goetz, E. T. (1991). A critique of schema
theory in reading and a dual coding alternative. Reading Research
Quarterly, 26, 463-484.

Salyer, M. G. (1990). The significance of difficult vocabulary to
reading in a second language. Unpublished Doctoral Dissertation.
Michigan State University.

Samuels, S. J. (1977). Introduction to theoretical models of readings.
In W. Otto (Eds), Reading problems Boston: Addison Wesley.

Samuels, S. J., 8 Farstrup, A. E. (1992). What research has to say
about reading instruction. Newark, DE: lntemational Reading
Association.

Samuels, S. J., 8 Kamil, M. L. (1988). Models of the reading process.
In P. L. Carrell, J. Devine, 8 D. E. Eskey (Eds), interactive approaches
to second language reading. New York: Cambridge University Press.

Samuels, S. J. 8 LaBerge, D. (1983). A critique of a theory of
automaticity in reading: Looking back: A retrospective analysis of
the LaBerge-Samuels reading model. In L Gentile, M. Kamil 8 J.
Blanchard (Eds), Reading research revisited. Columbus, Ohio: Charles
E. Merrill Publishing Company.

Sarig, G. (1987). High-level reading in the first and in the foreign
language: Some comparative process data. In J. Devine, P. L. Carrell,
8 D. E. Eskey (Eds), Research in reading in English as a second
language. Washington, D. C.: TESOL.

Schraw, G., 8 Sperling Denninson, R. (1993). The effect of reader
purpose on interest and recall. Joumai of Reading Behavior, 26, 1-
18. -

Schulz, R. A. (1983). From word to meaning: Foreign language reading
Instruction after the elemmentary course. The Modern Language
Joumai, 67, 127-134.

170

Schwanenflugel, P. J. (1991). The psychology of word meanings.
Hillsdale, NJ: Erlbaum Associate, Publishers.

Schwanenflugel, P. J. (1991). Why are abstracts concepts hard to
understand? In P. J. Schwanenflugel (Eds), The psychology of word
meanings. Hillsdale, NJ: Erlbaum Associate, Publishers.

Shimoda, T. (1993). The effects of interesting examples and topic
familiarity on text comprehension, attention, and reading speed.
Journal of Experimental Education, 61, 93-103.

Simonsen, 8., 8 Singer, H. (1992). Improving reading instruction. In S.
J. Samuels 8 A. E. Farstrup (Eds), What research has to say about
reading instruction. Newark, DE: lntemational Reading Association.

Smith, F. (1982). Understanding reading. (3 rd. Ed.). New York: Holt,
Rinehart 8 Winston.

Stahl, S. A. (1991). Beyond the instnlmental hypothesis: Some
relationships between word meanings and comprehension. In P. J.
Schwanenflugel (Eds), The psychology of word meanings. Hillsdale,
NJ: Erlbaum Associate, Publishers.

Stahl, S. A., 8 Chour Hare, V. (1991). Defining the role of prior
knowledge and vocabulary in reading comprehension: The retiring of
number 41. Joumai of Reading Behavior, 23, 487-508.

Stahl, S. A., Hare, V. C., Sinatra, R., 8 Gregory, J. F. (1991). Defining
the role of prior knowtedge In reading comprehension: The retiring of
number 41. Joumai of Reading Behavior, 23 (4), 487-508.

Stahl, S. A., 8 Jacobson, M. G. (1986). Vocabulary knowledge, prior
knowledge, and text comprehension. Joumai of Reading Behavior, 18,
309-323.

Stahl, S. A., Jacobson, M. G., Davis, C. E., 8 Davis, R. L. (1989). Prior
knowledge and difficult vocabulary in the comprehension of
unfamiliar Text. Reading Research Quarterly, 24, 27-43.

171

Stanovic, K. (1980). Toward and interactive-compensatory model of
individual differences in the development of reading fluency. Reading
Research Quarterly, 16 (1 ), 32-71.

Steffensen, M. S. (1987). The effect of context and culture on
children's L2 readlng: A review. In J. Devine, P. L. Carrell, 8 D. E.
Eskey (Eds), Research in reading in English as a second language.
Washington, D. C.: TESOL.

Stein, B. 8., Brock, K. F., Ballard, D. R., 8 Vye, N. J. (1987).
Constraints on effective pictorial and verbal elaboration. Memory 8
Cognition, 15, 281-290.

Stevens, K. (1980). The effect of topic interest on the reading
comprehension of higher ability students. Joumai of Educational
Research, 73, 365-368.

Sticht, T. G. 8 McDonald, B. A. (1992). Teaching adults to read. In S. J.
Samuels, 8 A. E. Farstrup (Eds), What research has to say about
reading instruction. Newark, DE: lntemational Reading Association.

Stoller, F., 8 Grabe, W. (1993) Implications for L2 vocabulary
acquisition and instruction from L1 vocabulary research. In Huckin,
T., Haynes, M., 8 Coady, J. (Eds), Second language reading and
vocabulary leamlng. New Jersey: Ablex Publishing Corporation.

Strother, J. B., 8 Ulijn, J. M. (1987). Does syntactic rewriting affect
English for science and technology text comprehension?. In J. Devine,
P. L. Carrell, 8 D. E. Eskey (Eds), Research in reading in English as a

second language. Washington, D. C.: TESOL.

Swaffar, J. K. (1988). Readers, texts, and second languages: The
interactive processes. The Modern Language Joumai, 72, 123-149.

Taylor, B. M. (1992). Text structure, comprehension and recall. In S.
J. Samuels, 8 A. E. Farstrup (Eds), What research has to say about
reading instruction. Newark, DE: lntemational Reading Association.

Tierney, R. J., Readence, J. E., 8 Dishner, E. K. Reading strategies and
practices. Needham Heights, MA: Allyn and Bacon.