: . . . . . V . . . .. I . V , . V .
. , . , .. A . ,

v , .. Z . V . . V . 1. AV A.

. . V V . . . ,. V. . V‘ . . .

. . V V, .f V V V , . .

n V V . V . . . A V . V

. .. . V. V . .. a, V\

~ ‘ . .., . . ,. ..
L. V. V .. V . . V . V . . ..

.. . V V . . .. . . . v f: 1. . V .c:

. . A V V ‘

. , .

 

N361 @051 9\

m Illlllllllllllllllllll'vllllllljllII
r1 ,

3 1293 00563 32(‘)|
LIBRARY
Michigan State

EUnivenity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled

The Effects of the Self-concept on the

Performance Appraisal Process

presented by

David Brent McKellin

has been accepted towards fulfillment
of the requirements for

M.A. degree in Psychology

\ %\
Major professorfl

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

Date 51121 89

 

 

 

MSU

 

 

RETURNING MATERIALS:
Place in book drop to

 

 

LIBRARIES remove this checkout from
-;—-.. your record. FINES will
be charged if book is
returned after the date
stamped below.
$362€ . " ""

,5;

33.159? 1 lif-‘Q'S

7... ‘l l

 

 

THE EFFECTS OF THE SELF-CONCEPT ON THE
PERFORMANCE APPRAISAL PROCESS

By
David Brent McKellin

A THESIS

Submitted to
Michigan State University
in partial fulfillments of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1989

,---_. .. (_ , .
WU, )(CC'?

ABSTRACT

THE EFFECTS OF THE SELF-CONCEPT ON THE
PERFORMANCE APPRAISAL PROCESS

By
David Brent McKellin

The relationship between one’s self-concept, defined as a set of cognitive
generalizations about the self, and the search for performance information about
another person was examined. Using a computerized information display board to
present information, 101 undergraduate psychology students, all former employees
in a fast food restaurant, searched for information about an hypothetical
subordinate. While only one marginally significant relationship was found between
a rater’s self-concept and his or her search for information, significant relationships
were found among a rater’s self-rated performance, experience performing the tasks
being rated, and ratings of task’s importance for successful job performance.

Limitations of the study and suggestions for future research were discussed.

ACKNOWLEDGMENTS

At long last, it’s finally done! Like any project worth doing, this thesis has
required a great deal of blood, sweat, and tears, but it’s all been worth it. Therefore,
I want to take this opportunity to express my heartfelt thanks to all those friends
who helped me survive this process.

First, I’d like to thank the members of my committee, Dan Ilgen, Kevin Ford,
and Steve Kozlowski for their encouragement and very valuable input on this
project. Their suggestions and insights were integral in making this research work,
and I am very appreciative of their help. I am especially grateful to Dan for
agreeing to chair my committee. I’ve learned a great deal working with him these
last couple of years, and have really appreciated his numerous recommendations,
constant support, and the time he has spent reading countless numbers of drafts of
this thesis. His feedback has been invaluable in helping me think about research in
a structured way and in expressing my ideas in a logical, coherent manner. Thank
you!

In addition to my committee, I’d like to thank John Hollenbeck for helping
me figure out how to analyze my data. I didn’t really believe it at the time, but all of
that preaching about canonical correlations in MGT 906 really paid off! I’d also like
to thank the Baker Basement crew for all of their help and support. Special thanks

go to Sue Schechtman and Mary Doherty for making the process tracing program

iii

available to me and for helping me figure out how it works. Also, thanks to Jeff,
Mary, MaryBeth, and Scott for being there when I needed to blow off steam as well
as when things were going well.

I’d also like to express my appreciation to my undergraduate help, Julie and
Keenan, who assisted me in data collection and keypunching. Thanks as well to
McDonalds and Burger King restaurants, whose assistance in creating the stimulus
materials for this project was crucial.

Finally, thanks to my family for their remarkable support. To my brothers,
Bill and Dan, thanks for your encouragement and for at least trying to appear
interested in what I’ve been studying. To my parents, Bill and Ruth McKellin, my
love and thanks for giving me life, for helping me develop a healthy curiosity into
how things work, and for your constant love. Dad, I wish you were still here to help
celebrate this accomplishment. Finally (and most importantly), to my incredible
wife, Ruth Ann, your patience, support, and love are beyond belief! Thank you for
putting up with me when I was frustrated, celebrating with me when I reached a
milestone, for proofing hundreds of pages of drafts, and for everything else that has
gone with finishing this thesis. You deserve this degree every bit as much as I do. I

love you!

iv

TABLE OF CONTENTS

LIST OF TABLES .................................................... viii
LIST OF FIGURES ................................................... ix
INTRODUCTION .................................................... 1
Information Acquisition in Performance Evaluation .................. 2

The present study ......................................... 5

Purpose and Overview ........................................... 6
Cognitive Concepts Relevant to Performance Appraisal .............. 6
Categories and Schema .................................... 7

Individual Differences in Schema Usage and Perception ........ 8

Differences in which schemas become accessible ......... 9

Factors affecting schema accessibility .................. 10

Summary .......................................... 14

The Effects of Self-Concept on the Perception of Others .............. 15
Current theory on the structure of the self-concept ............. 15

The effects of the self-concept in person perception ............ 17

Selection of information about others .................. 18

Self-image bias in person perception ................... 20

Effects of expertise on person perception ............... 24

METHOD ........................................................... 29

Overview ...................................................... 29
Sample ........................................................ 29
Task .. ......................................................... 30
Procedures .................................................... 33
Questionnaire completion .................................. 33
Computerized information search task ....................... 34
Introduction and practice ............................ 34

Introduction to the information search task ............. 35

Search task ........................................ 35

Performance stimuli ................................. 35
Questionnaire measures ......................................... 37
Self-rated performance .................................... 37
Dimension importance ratings .............................. 37
Experience measures ...................................... 37
Demographic questionnaire ...... . ......................... 38

Other measures ................................................ 39
Search: Amount .......................................... 39
Search: Content .......................................... 39
RESULTS ........................................................... 40
Scale properties ................................................ 40
Control variables ............................................... 44
Descriptive data ................................................ 47
Hypothesis One ................................................ 54

Hypothesis Two ................................................ 57

Hypothesis Three ............................................... 57
Hypothesis Four ................................................ 61
Hypothesis Five ................................................ 65
Relative Experience ....................................... 65

Percent Worked .......................................... 69

Dimensional Analyses ..................................... 69

Post Hoc Analysis ............................................... 74
DISCUSSION ........................................................ 75
Summary of Results ............................................. 76
Limitations of the study .......................................... 78
Implications of the current study .................................. 81
Directions for future research ..................................... 82
Conclusion .................................................... 85

LIST OF REFERENCES .............................................. 87
APPENDIX A - Consent Form ......................................... 94
APPENDIX B - Background Questionnaire ............................... 95
APPENDIX C - Computer Task Tutorial ................................. 98
APPENDIX D - Computer Search Practice Task .......................... 115
APPENDIX E - Computer Task Introduction Screen ....................... 119
APPENDIX F - Computer Task Behavioral Descriptions ................... 120
APPENDIX G - Self-rated Performance Questionnaire ..................... 123
APPENDIX H - Dimension Importance Rating Questionnaire ............... 127
APPENDIX I - Zero-Order Correlations Among Measures .................. 131

vii

LIST OF TABLES

Table 1. Internal Consistency Reliabilities (Coefficient alpha) for Self-rating

and Dimension Importance Scales ................................. 41
Table 2. Correlations among Self-rated Performance Scales ................. 42
Table 3. Correlations among Dimension Importance Rating Scales ........... 43
Table 4. Canonical Correlations Among Primary Research Factors ........... 46
Table 5. Partial Correlations Among Primary Research Variables ............ 48

Table 6. Canonical and Bivariate Correlations and Redundancy Index
Between Self-ratings and Task Importance .......................... 55

Table 7. Canonical and Bivariate Correlations Between Self-ratings and
Information Search ............................................. 58

Table 8. Relationship Between Relative Experience and One Piece
Partialling Importance ........................................... 64

Table 9. Relationships Among Relative Experience, Self-ratings, and
Importance .................................................... 68

Table 10. Relationships Among Percent Worked, Self-ratings, and
Importance .................................................... 7O

viii

LIST OF FIGURES

Figure 1. Relationship between Self-evaluated Performance and Perceived
Dimension Importance. .......................................... 22

Figure 2. Relationship between Self-evaluated Performance and Information
Search. ........................................................ 23

Figure 3. Perceived Dimension Importance as a Mediator Between
Self-evaluated Performance and Information Search. ................. 24

Figure 4. Perceived Dimension Importance as a Mediator between

Experience and Information Search. ............................... 27
Figure 5. Relationships among Experience, Self-evaluated Performance,

and Perceived Dimension Importance .............................. 28
Figure 6. Job Dimension and Behavioral Alternative Selection Screen ........ 31
Figure 7. Sample of Employee Performance Description. ................... 33

Figure 8. Canonical Correlations Among Self-ratings, Importance, and
Information Search. ............................................. 60

Figure 9. Canonical Correlations Among Experience, Importance, and
Information Search ............................................. 63

Figure 10. Canonical Correlations Among Experience, Self-ratings, and
Importance. .................................................... 66

Figure 11. Dimensional Relationships Among Experience, Self-Ratings and
Importance. .................................................... 71

INTRODUCTION

The evaluation of employee performance is perhaps one of the most
pervasive and important personnel issues that organizations confront. Defined as
the process by which an observer, usually a supervisor, rates the job performance of
an employee, performance appraisals are usually conducted annually or
semiannually (DeNisi, Cafferty, & Meglino, 1984; Ilgen & Feldman, 1983). The
information resulting from the appraisal may then be used for providing feedback to
employees regarding the quality of their achievements, as well as for human
resource decisions regarding promotion, compensation, training, and employee
development.

Since performance appraisals are so pervasive, a great deal of research has
been conducted on them. Landy and Farr (1980) reviewed and critiqued 30 years of
research on performance measurement in industrial settings. While they found that
an impressive number of studies used performance judgments as criteria, they also
acknowledged the susceptibility of these ratings to both intentional and
unintentional biases. Attempts to understand these biases often led to unsystematic
and atheoretical inquiries regarding the effects on ratings of (a) the context in which
the rating was completed (e.g., administrative purpose vs. research purpose; Heron,
1956), (b) rater and ratee personal characteristics (e.g. sex, race; Hamner, Kim,

Baird, & Bigones, 1974), and most abundantly (c) rating scale format (e.g. Latham

2
& Wexley, 1977). Additional efforts to control biases in performance ratings
concentrated on rating errors by focusing on training raters to recognize and avoid
these errors. Observing that the direction performance appraisal research was
proceeding appeared to be leading to diminishing returns, Landy and Farr (1980)
suggested that the emphasis of performance rating research shift and concentrate
instead on the rating process as a whole. In particular, they recommended that
special emphasis be placed on the cognitive processes of the rater.

In response to the cognitive processing suggestion, several models of rater
cognitive processes have appeared (e.g., DeNisi, Cafferty, & Meglino, 1984;
Feldman, 1981; Ilgen & Feldman, 1983). Central to these models is the view that a
rater’s task consists of making a performance judgment by processing information
about the ratee. More specifically, the judgment process is posited to proceed
through four ordered cognitive processes: (a) attention to and observation of
performance information, (b) encoding of this information and its storage in
memory, (c) retrieval of such information from memory prior to evaluation, and
finally, (d) integration of available information retrieved from memory in order to
make a rating judgment. These four are often presented as four stages and labeled
acquisition, storage, retrieval, and judgment.

Information Aoouisition io Porfogognoo Evoloatioo

While rater cognitive processing has dominated the performance appraisal
literature since 1980 (Ilgen, Barnes-Farrell, & McKellin, 1986), the four stages have
not received equal attention. In particular, the information acquisition stage of the
process has not received much study in the appraisal literature. This is unfortunate

since job behaviors must first be noticed before they can be stored in memory for

3
later recall and evaluation. Thus, all later stages hinge on the adequacy of the initial
phase.

Within the acquisition literature, the primary emphasis has been on
characteristics of the rating setting, specifically on the rating purpose and its effects
on how information is selected about others’ performance. In studying how the
purpose of the appraisal affected the extent to which subjects search for information
about ratees, Matte (1982) argued that raters would consider using ratings for
making promotions more important than using them to give ratees feedback. He
then hypothesized that raters would search for more information about a ratee’s
performance when they believed the rating task was more important than when they
believed it was less important. While Matte was unable to establish that raters
actually believed one purpose was more important than the other, his results did
indicate that raters searched for information more extensively when the appraisal
was to be used for both a promotion decision and feedback than when it was for
feedback only.

Schechtman (1987) also examined the effects of rating purpose on raters’
information acquisition. In this study, raters were instructed to search for
information about several employees for the purpose of either selecting an
employee to promote (administrative purpose) or selecting an employee to receive
remedial training (developmental purpose). She found that the purpose of the
rating affected how much information raters searched for about ratee performance,
but only in conjunction with the ratee’s performance level. The highest level of
search occurred for good performing employees under promotion conditions and

poor performing ones when the purpose was to administer remedial training. She

 

 

4
interpreted the pattern of these results to indicate that rating purpose did influence
raters’ information acquisition strategies, but only when taking the ratee’s
performance level into account.

Finally, Williams, DeNisi, Blencoe, & Cafferty (1985) investigated how
performance appraisal purpose and the type of decision to be made affected the
types of information raters search for about ratees’ performance. In this study,
raters were instructed to search for performance information about four individuals’
performance. The information which was available took three forms: (a) how well
the ratee performs a given task on other occasions (consistency information), (b)
how well the ratee performs on other tasks within his or her job scope 3
(distinctiveness information), and (c) how well the other three workers perform on
the same set of tasks (consensus information). The raters’ task was to search for
information about the ratees’ performance for the purpose of determining salary
increases, promotions, or remedial training needs. Using this information, raters
were told to either determine how oesorving the ratee was for a particular type of
action (e.g. promotion or training), or to actually designate that the ratee should
receive the specified outcome. Williams, et. a1 did not find relationships between
either appraisal purpose or the type of outcome decision to be made and
information search. They did, however, find a significant interaction between the
rating purpose and the type of outcome (i.e., designation decision or deservedness
rating) resulting from the rating. While similar information request patterns were
found across all three rating purposes, subjects who rated how deserving the ratee

was of a salary increase searched for distinctiveness information most, consistency

second, and consensus least. Ratees who designated whether or not the ratee

5

should receive a salary increase, on the other hand, requested consensus
information most, followed by distinctiveness and consistency information.

Research on the appraisal context, then, has found that rating purpose
interacts with other contextual factors to affect how raters select information about
others’ performance. The specific nature of this effect, however, has not been
thoroughly investigated, and therefore is not completely clear at this time.
W

The classic model for the perception of any stimuli is that perception is a
function of characteristics of the environment or context and those of the person
whose perceptions are of interest. While rating context has been found to affect
raters’ information acquisition patterns, no literature in the performance appraisal
area has investigated how Wages among raters affect the information
acquisition process. However, there is literature in the area of social cognition that
has examined individual differences in perceptions of others. One characteristic
which has been found in that literature to affect attention to others is the observer’s
self-concept (e.g. Pong & Markus, 1982). The social cognition literature defines the
self-concept as a set of cognitive generalizations containing information about the
self (Markus, 1977). Specifically, social psychologists have found that people’s self-
concepts affect the traits to which they attend when describing others’ personalities
(e.g. Fong & Markus, 1982; Lewicki, 1983, 1984).

The present research explores the effects of the self-concept on the cues to
which raters attend when doing a performance appraisal. However, for
performance appraisal, the criteria have been ratee behaviors, not personality traits.

Although it will be argued that self-concept effects on attention to ratee behaviors

 

6
are likely to be similar to those observed for traits, it is not clear at this point that
this generalization is to be expected. Because of differences in constructs and
operational definitions between trait- and behaviorally-based research, it is
necessary to conduct research with the explicit purpose of focusing on performance
appraisals to judge whether such an effect is relevant to the appraisal setting (Ilgen,
Barnes-Farrell, & McKellin, 1986; Ilgen & Favero, 1985). With this in mind, the
present research investigated the possibility that a rater’s self-concept, an individual
difference variable, affects how he or she searches for information about others’
performance.
W
To understand how individual differences in raters’ cognitive systems affect
the information to which they attend concerning others’ performance, this paper
proceeds as follows. First, basic concepts from cognitive psychology which relate to
individual differences in information storage will be discussed along with the effects
of storage processes on the attention to performance cues. Next, the role that one
specific individual difference variable, self-concept, plays in the perception of social
cues about others will be discussed. Building on this background, it is then proposed
that raters’ self-concepts affect the set of cues to which they attend in judging the
performance of others.
E . . C B l E E E . I
To understand how a person’s self-concept affects the processing of
information about others, it is first necessary to review some basic concepts related
to how information is stored in memory. The discussion that follows briefly

describes two organizing systems, categories and schemas, which are thought to

 

7
affect how information is processed and stored in memory. Next, individual
differences in how a specific category or schema is selected for processing
information is examined. With an understanding of how these organizing systems
are selected, it is then possible to see how this selection is thought to affect the
cognitive performance appraisal process. Finally, how a person’s self-concept can
affect the process by which information about others’ performance is selected from
one’s social environment is addressed.
Categoriesandfichema

Borrowing heavily from research in social cognition and cognitive psychology,
current performance appraisal theory is based on the assumption that raters
systematically encode ratees’ routine behaviors in memory (Feldman, 1981; Ilgen &
Feldman, 1983; Shiffrin & Schneider, 1977). The cognitive literatures assume that
encoding is achieved by utilizing cognitive structures, typically labeled either
categories or schemas. Categories and schemas function in memory to represent
organized knowledge about a given stimulus (Bartlett, 1932; Lord & F oti, 1986).
The term ”category" is used most frequently to refer to a "fuzzy set" of behaviors
which share similar features. The prominent features reflect typical category
characteristics (Rosch, 1978). The term schema is used to generally refer to highly
structured organizational systems of pre-existing information. Using these systems,
incoming information is stored according to its serial order of occurrence, shared
characteristics or elements, or relevance to a particular person (Ilgen & Feldman,
1983; Lord & Foti, 1986). In other words, a category is a presumed set or "bin" and
the schema is the system that establishes that bin. In the literature, the two are

often used as if synonymous. Functionally, both are believed to serve as guides for

8
the interpretation of incoming information and to assist in the subsequent
generation of appropriate actions and expectations (Graesser, Woll, Kowalski, and
Smith, 1980; Lord & Foti, 1986; Taylor and Crocker, 1981).
ivi Di n ' h m n P i It

Previous research has shown that people differ in the number and the nature
of their schemas and categories (Feldman, 1981; Markus, 1977; Sechrest, 1968).
These differences may be due to cultural factors (T riandis, 1976), prejudice,
cognitive complexity (Feldman & Hilterman, 1975), or expertise (Markus, Smith, &
Moreland, 1985), to name a few possible influencing factors, and may affect the
salience of different schemas for different people. There are also differences in the
likelihood that an individual will use any particular schema to process information
(Higgins, King, & Mavin, 1982; Tversky & Kahnemann, 1973, 1974; Srull & Wyer,
1979; Wyer & Srull, 1980). Specifically, one factor thought to influence the
likelihood that a schema will be used in information processing is the accessibility of
that schema for use in memory encoding (Higgins & King, 1981; Higgins, King, &
Mavin, 1982; Wyer & Srull, 1980, 1981).

Two different approaches have been used to investigate the effects of
heightened schema accessibility on information processing. The first of these
approaches, which focuses on the effects of situational variables on accessibility,
typically has involved "priming" the schema of interest. Concentrating on short-term
accessibility, priming refers to the effect of presenting cues in one context on
increasing a schema’s accessibility in a subsequent context (Fiske & Taylor, 1984).

While priming studies have concentrated on situational effects on short-term

accessibility, there may also be differences among individuals that cause schemas to

9
become accessible for longer periods of time (Wyer & Srull, 1980). In fact, if a
schema continues to be used on a regular, frequent basis, it is said to become
"chronically accessible,” meaning that its availability and selection for use in
information processing is unusually high when compared to other schemas that
might be implemented (Higgins, Rholes, & Jones, 1977; Srull & Wyer, 1979, 1986;
Wyer & Srull, 1980). Therefore, if a schema becomes chronically accessible it will
be employed as though it is constantly primed, and it may become a stable
characteristic of the individual’s information processing system (Bargh & Thein,
1985; Higgins & King, 1981; Higgins, King, & Mavin, 1982).

The next section reviews a portion of the research which has studied the
effects of chronic schema accessibility on memory encoding and its subsequent
effects on the salience of environmental cues. This review provides the background
for addressing how the concept of schema accessibility relates to the process of
making performance judgments.

WM. Which schemas become
chronically accessible varies across individuals. For example, Higgins, King, and
Mavin (1982) studied the extent to which people differ in their accessible schemas in
a two-part study. In the first part, subjects were asked to list the traits of (a) a type
of person they liked, (b) a type of person that they disliked, (c) a type of person they
sought out, (d) a type of person they avoided, and (e) a type of person they
frequently encountered. Approximately one week after completing this task,
subjects participated in a second, "unrelated" study in which each read an
individually tailored essay containing 12 behavioral descriptions of a target person.

These descriptions had been constructed such that the 6 traits the subject had listed

10

first during Part One of the experiment ("accessible traits") were embedded in the
essay the subject read, as were 6 traits included in lists compiled by other subjects,
but which the subject had not included in his or her own lists ("inaccessible traits").
In order to assess the effects of trait accessibility on information processing, Higgins
et al. administered a 10-minute nonverbal task immediately after subjects read their
essays, and then asked subjects to exactly reproduce the essay from memory. In
addition, subjects were also asked to describe the target person as fully as possible.

Results of this study demonstrated two important characteristics of chronic
schema accessibility. First, Higgins et al. found that 50 of the 88 distinct traits
determined to be accessible in the initial trait listing task (57%) were listed by only
one subject. In addition, the trait found to be most accessible ("Humorous") was
actually found to be accessible for only 38 percent of the subjects. These results led
the authors to conclude that there appears to be relatively little overlap in people’s
accessible schemas. Second, when reproducing the essays, subjects recalled fewer
aspects of the description associated with their inaccessible schemas than details
associated with their accessible schemas. These findings suggest that schema
accessibility, especially when a schema is chronically accessible, can be construed as
an individual difference characteristic that varies widely across people.

antors affooting sohoma aooossioility. Current concerns. Schemas may
become chronically accessible through a variety of means. First, an individual may
be sensitized toward using a particular schema because of his or her "current
concerns" (Klinger, 1975); that is, the interaction of a person’s immediate needs and
values with stimuli in his or her environment may affect which schema is accessed

(Srull & Wyer, 1986). No known empirical studies have investigated the specific

11
effect that current concerns may have on a schema’s accessibility for encoding social
information. However, in studies such as that by Luria and Vinogradova (1959),
subjects were found to experience increased physiological arousal (an increase in
finger and scalp blood volume) when a word for which they were instructed to listen,
or one of its synonyms, was presented in a list. This suggests that an individual’s
current concerns (listening for the target word) increase sensitivity to stimuli related
to those concerns (Klinger, 1975). This type of effect might be typified in industrial
and organizational psychology when people interview for jobs. Because job
candidates are so concerned with making the best impression, they tend to think of
every action or statement they make, or is made towards them, in terms of its impact
on their prospects for receiving a job offer. Thus, interviewees’ current concerns
may affect the accessibility of their "good job candidate" schema.

Personal values. Personal values may also influence schema accessibility. In
two studies by Bruner (1951, 1983), values were found to affect schema accessibility
in the perception of both ambiguous and unambiguous stimuli. Bruner (1951)
measured subjects’ value orientations using the Allport-Vernon Study of Values test.
Next, he used a tachistoscope to present a series of pictures, each corresponding to
one of six values (religious, economic, theoretical, social, political, and aesthetic),
and subjects were asked to describe what they saw. Each picture could be
interpreted as depicting one of the listed values or as being truly ambiguous. For
instance, one picture was of man bending over; it could be interpreted as a man
working (economic value) or praying (religious value). Results confirmed the
hypothesis that subjects’ value orientations affected their interpretation of what they

saw.

12

Similarly, the effects of personal values on schema accessibility were shown
to be extremely persistent in a study by Erdelyi & Appelbaum (1983). Drawing a
sample from members in a Jewish organization, Erdelyi & Appelbaum used a
tachistoscope to present subjects with an array of stimuli consisting of eight neutral
figures arranged in a circle. In the middle of this array, the authors introduced
another symbol, which varied across conditions. The ninth symbol was either a
swastika, a Star of David, or a window. Prior to viewing the array, subjects were
instructed that they should concentrate on remembering what figures were shown,
and their locations in the array. Next, subjects were shown the array, and were
asked to report what figures they had perceived and their locations. Results
demonstrated that the symbol presented in the middle of the array affected subjects’
ability to recall the other symbols in the array. Subjects who viewed the array with
the swastika correctly recalled significantly fewer symbols than those whose array
included the Star of David, who recalled fewer symbols than those who were shown
the window figure. This experiment suggests that long-term personal values can
influence information processing by directing people’s attention towards cues
associated with their values, even when concentration on such value-laden
interpretations is not completely appropriate. Overall, these studies demonstrate
that personal values can affect schema accessibility, thereby biasing information
processing in favor of schemas associated with one’s strongly held values.

Personality traits. Finally, stable personality traits have also been shown to
affect schema accessibility. Studies by Markus and others (e.g. F ong & Markus,
1982; Lewicki, 1983; Markus, 1977; Markus & Smith, & Moreland, 1985) have

investigated the information processing characteristics of individuals who describe

13
themselves as being near the extreme end of a particular trait dimension (possess
the trait or do not possess the trait) and who feel this domain is important for their
view of themselves. Research into the cognitive performance of these "schematics"
(Fong & Markus, 1982) has generally supported the hypotheses that they (a) process
information that is interpretable in terms of that dimension more quickly and (b)
remember it better (e.g. Bargh, 1982; Bargh & Pietromonaco, 1982; Markus, 1977;
Markus, Crane, Bernstein, & Siladi, 1982).

Besides increasing sensitivity to environmental stimuli, chronically accessible
personality schemas may also bias how some types of information are processed. In
the first stage of a two-part study on the biasing effects of personal traits, Markus
(1977) administered measures of the bipolar trait "Independence/ Dependence" to a
class of introductory psychology students. These measures included subjects’ self-
ratings of their own independence on the Gough-Heilbrun Adjective Check List
(Gough & Heilbrun, 1965), and on semantic differential scales behaviorally
describing the dimension Independent-Dependent. On the latter measure, subjects
were also asked to rate the importance of each semantic dimension for describing
themselves. From these measures, subjects who (a) rated themselves highly on at
least two of the scales (8-11 on an 11-point scale), (b) rated these dimensions
important (8-11 on an 11-point scale) and (c) checked themselves as "independent"
on the adjective check list were termed "Independents." Similarly, subjects who (a)
rated themselves at the low end of at least two of the three scales (1-4 on an 11-
point scale), (b) rated these dimensions as being important (8-11 on an 11-point
scale), and (c) checked themselves as "dependent" on the adjective check list were

identified as "Dependents." Finally, individuals who (a) rated themselves in the

14
middle range (5-7 on an 11-point scale) on at least two of the three scales, (b) fell in
the lower portion of the distribution on the importance scale, and (c) did not check
themselves as either "independent" or "dependent" on the adjective check list were
termed "Aschematics." This meant that subjects in this category did not possess
schemas relevant to the dimension Independent/ Dependent. Using this
information about subjects’ schematic development regarding Independence /
Dependence, 16 subjects from each category were selected to participate in the
main phase of the study.

In the main study, Markus studied the effect that being schematic for
Independence had on subjects’ reactions to false feedback concerning their
disposition on the similar trait "suggestibility." Measuring subjects’ acceptance of
this feedback, she found that schematics’ were less likely than aschematics to accept
incongruent or counterschematic information as self-diagnostic because it was
inconsistent with, and therefore threatening to, their own self-images. Therefore,
this study suggested that being schematic related to a particular trait dimension
might bias one’s processing of self-relevant information. Additional implications for
being schematic in certain domains will be discussed later in this paper.

Snanary. As evidenced by the preceding studies, it appears that schema
accessibility biases attentional processes towards schema-relevant information (cf.
Bargh & Pietromonaco, 1982; Geller & Shaver, 1976). Chronically accessible
schemas, it is assumed, become highly accessible to an individual because of his or
her (a) current concerns, (b) personal values, and/ or (c) stable personality traits. In
general, it appears that components of an individual’s identity, of which current

concerns, personal values, and personality traits are a part, help to explain

15
individual differences in which schemas become chronically accessible. With this in
mind, it is argued that information related to one’s self may affect the chronic
accessibility of the schemas used to process information about others. Specifically,
this research seeks to investigate if an individual’s self-concept, construed as a
collection of schemas which contain information about the self (Markus, 1977),
resembles a chronically accessible schema in its effects on how information is
selected about others whose performance is to be rated. In order to generalize this
effect to the performance appraisal context, the following section will review
research showing the effects of one’s self-concept on the selection and processing of
trait-related information about others.
Tthf flf-n nhPr i r

Psychologists since Freud have recognized that, through the phenomenon of
projection, one’s own personality characteristics are often attributed to others (e.g.,
Freud 1924/1956; Holmes, 1978; Schiffenbauer, 1974). Even earlier theorists
recognized that one’s self-concept influences the perception of others (e.g., Hall,
1898; James, 1915). Mead (1934) Rogers (1951) and others (e.g. Sullivan, 1953)
have reached similar conclusions. But how a monolithic, unitary entity (which the
self-concept initially was perceived to be) was able to regulate cognitive processes in
a wide variety of contexts with varied outcomes became problematic within
established definitions of the self-concept. Therefore, it became clear to
contemporary theorists that an alternative view of the self was necessary.
Wm

One of the most important changes in the last decade of research on the self-

concept can be found in literature pertaining to its structure and content (Markus &

16
Wurf, 1987). A major impediment in linking the self-concept to the diversity of
behavior with which it was supposedly related was the view that very complex global
behavior was related to a stable, generalized, average view of the self, most closely
approximated by self-confidence. A solution to this problem has been found by
defining the self-concept as a set or collection of images, schemas, conceptions,
prototypes or theories about the self (Carver & Scheier, 1981; Epstein, 1980;
Greenwald, 1982; Kihlstrom & Cantor, 1984; Markus, 1983; Rogers, 1981;
Schlenker, 1980). Embracing this multidimensional, active view, Markus (1977)
defined the self-concept as a set of cognitive generalizations or representations
about the self, called self-schemas, acquired by gathering large amounts of
knowledge about the self in various areas through introspection. These
generalizations about the self serve to organize and guide the processing of self-
related information contained in the individual’s social experience. Using these self-
schemas, then, the self-concept aids the process of recognizing and interpreting self-
relevant stimuli in the various domains of an individual’s social environment.
Simply put, schemas correspond to how pe0ple perceive themselves.

Given the notion of the self-concept as a multidimensional construct, it was
no longer feasible to refer to the self-concept (Markus & Wurf, 1987). Rather,
because of limited cognitive capacity and the vast amount of information stored in a
person’s self-representations, the model of self-concept has come to describe the
working, on-line, or accessible self-concept (Cantor & Kihlstrom, 1986; Markus &
Nurius, 1986; Schlenker, 1985). This idea reflects the belief that not all self-schemas
that are part of the complete self-concept are accessible at any one time. Rather,

the set of schemas that is available at the moment is most frequently called the

17
"working" self-concept. The working self-concept is best viewed as a continually
active, shifting array of a person’s accessible self-knowledge.

Approaching the self-concept as a dynamic entity gains support on several
grounds. First, this view is consistent with assumptions suggested by the symbolic
interactionist perspective (Mead, 1934; Stryker, 1980) which contends that there is
not a fixed or static self-concept, but only a current self-concept constructed from
one’s social experiences. Second, the multidimensional view allows for a self-
concept that can be both stable and malleable. Core elements of the self may
remain relatively unchanged by variations in one’s social environment and may
remain chronically accessible to the working self-concept because of their extensive
elaboration and central role in defining the self (Higgins, King, & Mavin, 1982).
Nevertheless, changes in factors such as prevailing social contexts and individual
motivation may vary the accessibility of more volatile self-representations. Finally,
support for the dynamic view'of the self-concept follows from the growing body of
research reviewed earlier which suggests that individuals are heavily influenced in
all aspects of judgment, memory, and overt behavior by their currently accessible
pool of thoughts, attitudes, and beliefs about themselves (e.g. Higgins & King, 1981,
N isbett & Ross, 1980). Therefore, the notion of a dynamic self-concept helps
explain both stable and variable aspects of an individual’s self-concept.

f h lf- in r n r i

A large body of literature has been generated in the social cognition area
showing that the self-concept, through the use of self-schemas, is involved in person
perception (e.g. Bargh, 1982; Fong & Markus, 1982; Kuiper, 1981; Lewicki, 1983,
1984; Markus & Smith, 1981; Markus, Smith, & Moreland, 1985). All of this

18
literature will not be review here because the primary interest in the current study is
the self-concept’s role in selecting information about others. With this in mind, only
research on how the self-concept affects the selection of information is addressed.

W. A number of studies in social
psychology have demonstrated that people tend to use traits which they use to
define themselves when they search for information to describe others (e.g. Lewicki,
1983, 1984; Shrauger & Patterson, 1974). These studies, however, used methods
which only inferred that a person’s self-concept influenced this process. For
example, Shrauger and Patterson (1974) asked subjects to write several descriptions
of pe0ple they knew, and later to select 10 traits about which they were most
satisfied with themselves and 10 traits about which they were least satisfied with
themselves. Though they found that people tended to use dimensions which were
highly selfodescriptive when describing others, Shrauger and Patterson did not
directly establish a causal relationship between one’s self-concept and information
attended to in others.

In another study, F ong and Markus (1982) sought to determine if a person’s
self-concept directly affects the characteristics to which he or she attends in
another’s personality. Specifically, Fong and Markus investigated the impact of one
possible self-schema dimension, whether or not an observer had a self-schema for
introversion or extroversion, on the types of information he or she requested in
order to find out about others. In this study, subjects were identified as either
introvert schematics (those having self-schemas for introversion), extrovert
schematics (those having self-schemas for extroversion), or aschematic (those with

self-schemas that were not easily identified as introverted or extroverted). Next,

19
subjects were instructed to select 12 of a possible 26 questions they would ask to get
to know another person. This list included 11 questions judged to elicit information
about extraversion, 10 about introversion, and 5 irrelevant to either introversion or
extroversion. Results supported the hypothesis that people tend to seek information
about others that is related to dimensions for which they have self-schemas;
extrovert schematics selected more extrovert-related questions and introverts
selected more introverted questions. This study, along with those which used less
direct approaches (e.g. Lewicki, 1983, 1984; Shrauger & Patterson, 1974) provides
evidence that an individual’s self-schema related to a particular trait dimension
affects his or her selection of information about that dimension when evaluating
others.

Though the studies cited agree that people use self-relevant traits or
dimensions to judge others, exactly how self-schemas actually affect this process is
unclear. One perspective on how information is selected suggests that people are
biased to protect their self-images when judging others (Lewicki, 1983, 1984).
Another point of view, suggested by Kuiper and her associates (Kuiper, 1981;
Kuiper & Derry, 1981; Rogers, Kuiper, & Kirker, 1977), is that self-schemas affect
the selection of dimensions to observe because they, the self-schemas, serve as
anchors that observers find helpful in encoding their observations about other
people. This position is elaborated upon by Markus and her colleagues (e.g. Fong &
Markus, 1982; Markus, Smith, & Moreland, 1985) who suggest that people develop
self-schemas in those areas of their own personalities with which they most strongly
identify. These self-schemas, then, help people develop expertise related to these

domains. The degree of expertise subsequently affects the salience of information

20
about others. Therefore, from this point of view, the self serves as a reference point
which facilitates the development of expertise in various domains and in turn affects
which information a person attends to when observing others. For purposes of
discussion, I label the two perspectives just described as the "self-image bias" and the
"expertise view" and will discuss each below.

W- The self-image bias view of person
perception, presented by Lewicki (1983, 1984), suggests that when selecting
dimensions along which to evaluate others, an observer chooses those areas about
which he or she feels most positively about himself or herself. This, in turn,
enhances the probability that others will be seen as inferior if one assumes a
relatively normal distribution of perceived preference. In a study which investigated
subjects’ self-ratings and ratings of others, Lewicki (1983) instructed subjects to rate
themselves and 20 stimulus persons (who were known to them) along 25 trait
dimensions. Using these ratings, the centrality of each dimension was determined
by examining the amount of variance each rating dimension accounted for in all of
the other dimensions (i.e., by adding the squared product-moment correlations
between one dimension and each of the remaining 24 dimensions (Wishner, 1960)).
Results of this study showed a linear relationship between the desirability of an
individual’s self-rating on a trait dimension and its centrality or importance in
perceiving others. Lewicki hypothesized that this finding reflected a classic
motivational defense mechanism that protects a high self-evaluation, and called this
phenomenon the "self-image bias in person perception."

In a second study, Lewicki (1984) asked subjects to identify whether each of

four stimulus persons possessed 15 trait attributes (yes-no responses). Using

21
subjects’ response times as the dependent measure, Lewicki found that the
desirability of a person’s self-evaluation on a trait was linearly and inversely related
to the time it took them to identify whether the stimulus person possessed that trait.
This effect Lewicki attributed to the influence of self-schemas on the perceived
desirability of these traits, with more desirable traits being identified more quickly
than less desirable traits.

While these and other studies have demonstrated that the self-concept, as a
set of self-schemas, affects what information is noticed or considered important in
observing others in relation to trait dimensions, no literature has previously
investigated the role of the self-concept as it pertains to behaviorally-based domains.
By definition, the self-concept is a set of cognitive generalizations or schemas
containing information about the self. Though no research has addressed the issue,
this definition implies that the self-concept includes information about not only
trait-based information, but information about one’s own performance as well. The
inclusion of both behaviors and traits is consistent with the current behavioral
emphasis in the performance appraisal literature.

Accepting this more comprehensive construal of the self-concept construct
suggests that the self-concept should impact the processing of behaviorally-based
information in a way similar to the way it impacts the processing of trait-related
information. Therefore, when applying this definition within the framework of the
self-image bias perspective, one would expect that the favorability of a person’s
evaluation of his or her own performance on a behavioral dimension will be linearly
related to how important he or she perceives this dimension. Generalizing such an

effect to the performance rating context suggests that:

22

Hypothesisl: When presented with a set of behaviorally-based job
dimensions varying in perceived importance for successful job
performance, raters’ evaluations of those job dimensions’ importance
will be positively and linearly related to their evaluations of their own
performance on those dimensions (See Figure 1).1

 

 

 

Perceived
Self Wuamd. —) Dlmenslon
Performance Importance

 

 

 

 

 

Figure 1. Relationship between Self-evaluated Performance and Perceived
Dimension Importance.

The self-image bias perspective also suggests that observers attend more
readily to those characteristics of other people on which they (observers) evaluate
themselves most positively. If a similar effect were true in searching for
performance-based information about others, observers should search for more
information about others on those behavioral dimensions for which they evaluate
themselves most positively. Therefore, in the context of collecting information for
performance ratings, it is hypothesized that:

Hypothesis 2: When presented with a set of behaviorally-based job

dimensions, raters will search for more information about others’

performance on those behaviorally-based job dimensions for which

they evaluate their own performance most favorably than on those for

which they perceive their own performance to be least favorable (See
Figure 2).

 

‘ This figure and others that follow use directional arrows based upon the

implications behind the development of the hypotheses. However, it is recognized
that limitations in the methods used here cannot provide direct tests of causal
effects. The wording of the hypotheses is consistent with the causal limits of the
research.

23

 

 

Self-evaluated Information
Performance Search

 

 

 

 

 

 

Figure 2. Relationship between Self-evaluated Performance and Information
Search.

Hypothesis One addresses the effects of an individual’s assessment of his or
her own performance on perceived dimension importance, while Hypothesis Two
deals with the effects of his or her self-evaluated performance on the search for
information about others. It may be possible, however, to better explain the
relationship between self-evaluation, perceived importance, and information search
by considering all three factors simultaneously. Specifically, if both information
search and perceived dimension importance are related to one’s self-evaluation on
that dimension, it could be expected that dimension importance mediates the
relationship between one’s self-evaluation and one’s search for dimension-relevant
information about others. For dimension importance to mediate the relationship
between self-evaluation and information search implies two things: (1) there is a
relationship between how important one perceives a dimension to be and how much
information he or she searches for information about others’ performance on that
dimension; and (2) the strength of the relationship between self-evaluation and
information search changes in accordance with changes in the strength of the
relationship between dimension importance and information search. Such a

mediated relationship would mean that the relationship between self-evaluation and

24
information search would be relatively stronger (higher correlation) when the
relationship between importance and information search is stronger, and conversely
would not be evident when the correlation between importance and information
search is zero. With this relationship in mind, it is hypothesized that:

Hypothesifi: When presented with a set of behaviorally-based job
dimensions, the extent to which the favorability of raters’ evaluations
of their own performance on those job dimensions is related to the
amount of information they search for about others’ performance will
be mediated by how important those dimensions are perceived to be
for successful job performance (See Figure 3).

 

 

 

Self-evaluated P°'°°""d Information
Performance .mPort‘no. Search

 

 

 

 

 

 

 

 

 

Figure 3. Perceived Dimension Importance as a Mediator Between Self-evaluated
Performance and Information Search.

Effects of expertise on person pereeption. While Lewicki’s viewpoint implies

that a bias towards protecting one’s self-image affects the selection of dimensions
from which people gather information about others, Markus and her colleagues
(Fong & Markus, 1982; Markus, 1977; Markus & Smith, 1981; Markus, Smith, &
Moreland, 1985) hold the position that expertise is the key. This view suggests that
as a person develops self-schemas for the various important aspects of the self he or
she becomes particularly sensitive to stimuli related to these schemas. As time
passes, increased sensitivity to these schemas leads to the acquisition of large
knowledge bases in these areas, and the person develops expertise about them.

Markus and her colleagues have shown that, when dealing with information related

25
to domains for which an individual has developed self-schemas, he or she is able to
process this information more efficiently. In a study by Markus, Smith, & Moreland
(1985), subjects who had been classified as either schematic or aschematic on the
trait "masculinity" viewed a videotape of a person performing both neutral and
stereotypically masculine behaviors. They discovered that, when dealing with
information related to masculinity, schematics were able to observe the actions of
the person in the videotape and: (a) encode this information in memory using
larger units, or "chunks", of information; (b) draw a greater number of very
confident inferences from this information; and (c) respond more appropriately to
the specific observational purposes of the situation (e.g. general description of an
actor’s behavior vs. recall of specific behaviors) than were people who lacked self-
schemas in this domain. In addition, the sensitivity to schema-related information
suggested by this expertise may help explain the results of the F ong and Markus
(1982) study discussed previously in which people searched for more information
about others in domains corresponding to their self-schemas. Therefore, this
perspective asserts that people tend to concentrate on dimensions for which they
have developed the most expertise when judging others because these domains
provide them with more reference points for making decisions about others than
does attending to domains in which they do not consider themselves experts.

In attempting to generalize the effects of expertise to the selection of
performance information about others, it is necessary to determine what variables
might be. relevant to expertise. One relevant characteristic associated with expertise
is eggpetieooe in performing the task to be rated. It is expected that the more

experience a person has in performing a given task or job dimension, the more

26
expert he or she will be on that task. This expertise may affect the selection of
performance information about others in two ways. First, expertise may serve as a
frame of reference which could be used when evaluating the performance of others.
Across job dimensions, having a more developed frame of reference for some job
dimensions than for others might suggest that raters will search for more
information about a subordinate’s performance on job dimensions in which raters
have more experience than on those in which they have less experience.

A rater’s frame-of-reference for a job dimension may not have a direct effect
upon information search however. Rather, expertise concerning a job dimension
can also be thought to affect how important a dimension is perceived to be. In
addition to a more developed frame-of-reference, another outcome of expertise is
the ability to shift one’s focus from the details of actions to the "big picture"
(Markus, Smith, & Moreland, 1985). Thus, more experience in performing a
dimension should lead one to more fully understand how performing specific tasks is
important in meeting the overall goals. Therefore, it would appear that a job
dimension’s perceived importance would affect the relationship between a rater’s
experience on that dimension and his or her search for information about another’s
performance. Specifically, it is expected that the relationship between one’s
experience in performing a job dimension and searching for information about
another’s performance on that dimension changes with corresponding changes in
the perceived importance of that dimension. It is also expected that as the
correlation between dimension importance and information search approaches zero

the correlation between experience and information search will also approach zero,

27
indicating that dimension importance mediates the relationship between experience
and information search.

I_Iypotl;esis_4: When presented with a set of behaviorally-based job
dimensions, the extent to which raters’ experience on a job dimension
is related to their search for information about another’s performance
on that dimension will be mediated by how important they perceive
the dimension to be for successful job performance (See Figure 4).

 

 

 

 

 

 

 

 

 

P°'°°'V°d Information
Experlonoe p——>« Dlmonalon .—..
Importance Search

 

 

 

Figure 4. Perceived Dimension Importance as a Mediator between Experience and
Information Search.

Finally, the self-image bias and expertise explanations of the self-concept’s
effects on person perception may not be incompatible. Rather, these interpretations
of the self-concept’s effects may simply emphasize different aspects of the same
phenomenon. One similarity between the self-image bias and expertise perspectives
is that each view implies a direct influence on the importance one places on any
particular job dimension--self-image bias through self-evaluation and expertise
through job experience. When considering self-evaluation and experience together,
it becomes evident that one must first perform a task before evaluation on that task
can take place. Therefore, the self-image bias and expertise perspectives of person
perception may combine to explain perceived job dimension importance in that

experience may affect perceived importance indirectly through its relationship with

28
self-evaluation, as well as directly through its hypothesized umnediated relationship.

To determine if this is the case, it is hypothesized that:

Hypothesis 5: When presented with a set of behaviorally-based job
dimensions, raters’ experience in performing specific job dimensions
will be both directly related to their perceptions of a dimension’s
importance and indirectly related to importance through their
evaluations of their own performance (See Figure 5).

 

 

 

 

Exporlonc. \\
Poroolvod

 

 

 

 

 

1 Dlmonolon
Importanc-

Self-evaluated

Performance

 

 

 

Figure 5. Relationships among Experience, Self-evaluated Performance, and
Perceived Dimension Importance.

METHOD

Mew

The hypotheses were tested using a sample of raters varying in experience on
a job identical to the one performed by the persons to be rated. This was necessary
since raters’ work experience was to be related to their behavior when searching for
information about others’ performance on tasks they themselves had performed. In
addition to the search task, measures were collected regarding raters’ evaluations of
their own performance, of the importance of various job duties, and a number of
demographic variables. 1
Sample

The number of subjects required for this investigation was determined
through a power analysis (Cohen & Cohen, 1983). In the Fong and Markus (1982)
study reviewed earlier, self-schemas were found to account for approximately 21
percent (v_v2 = .21) of the variance in predicting which questions an observer selects
about another person. Desiring a power coefficient of at least .90, and taking into
account that two independent variables would be included in the analyses for
Hypotheses Three through Five (one variable without any prior variables, an
"exogenous" variable, and one mediating variable for each hypothesis), it was
determined that at least 52 subjects were required for this study. However, given

the exploratory nature of the study and the uncertainty about the extent to which the

29

30

Fong and Markus (1982) results would generalize from traits to behaviors, a sample
of 101 subjects was obtained (81 female, 20 male). All received course credit for
their participation. To insure that subjects understood the duties involved in the job
to be rated, participants were required to have experience working in one of three
fast food restaurants. This type of employment was selected primarily because of
the likelihood that members of the larger sample from which this sample was drawn
would have had this experience. Specifically, the sample was limited to current or
former employees of Burger King (n=25), McDonald’s (n=61), and Wendy’s
(n= 14) restaurants to control the types of tasks employees perform on the job.
Participants had a mean age of 18.83 years (SD = 1.35) and had worked in a fast
food restaurant for an average of 11.24 months (SD = 9.44). Eleven of the subjects
were currently employed by their restaurant, while 90 were not. Subjects no longer
employed by a restaurant terminated their employment an average of 19.56 months
prior to participating in this study (SD = 18.38).
Task

In order to record the amount of information for which subjects searched
within each job dimension, a computerized information display board was used.
Information display boards have previously been used to study information selection
strategies in both consumer psychology and other types of decision making research
(e.g. Doherty, 1987; Jacoby, Chestnut, Weigl, & Fischer, 1976; Payne, 1976). In
performance appraisal research, information display boards have been used
primarin to study the effects of various performance appraisal purposes on the
search for information about ratees (Matte, 1982; Schechtman, 1987; Williams,

DeNisi, Blencoe, & Cafferty, 1985).

31
The version of the information display board which was used in this study
was an interactive computer program. This program displays a list of job

dimensions and a list of behavioral alternatives from which subjects may select

 

information (See Figure 6).
DIMENSION ALTERNATIVE

1: RESTROOM MAINTENANCE 1: BEHAVIOR A
2: DINING ROOM MAINTENANCE 2: BEHAVIOR B
3: COUNTER SERVICE 3: BEHAVIOR C
4: SANDWICH PREPARATION 4: BEHAVIOR D
5: OUTSIDE MAINTENANCE 5: BEHAVIOR E
6: FRENCH FRY PRODUCTION 6: BEHAVIOR F

7: BEHAVIOR G

 

NTER NO. OF DIMENSION FROM 1 TO 6 THEN RETURN ?
NTER NO. OF ALTERNATIVE FROM 1 TO 7 THEN RETURN ?

 

 

 

Figure 6. Job Dimension and Behavioral Alternative Selection Screen.

Subjects’ task was to gather information about the performance of an hypothetical
employee. To do this, subjects were presented with a number of discrete choices.
First, subjects were asked to select which one of the possible six job dimensions on
which they would like information. Seven behavioral descriptions about the
employee’s performance were described within each of the six job dimensions.
After selecting a job dimension, subjects were next instructed to select which of
these descriptions they wished to display on the computer’s screen. Therefore, each

time a subject wished to display a piece of information about his or her hypothetical

32
employee, he or she would first select the job dimension from which that
information should come and then select which behavioral description within that
dimension should be displayed.

Actually selecting information to be displayed on the computer’s screen
involved entering information on the computer’s keyboard. To display an individual
piece of information, subjects first typed the number corresponding to the job
dimension about which they wanted information on the keyboard. Next, subjects
typed the number corresponding to the behavioral description they wished
displayed. Once both the dimension and alternative had been selected, the
computer displayed information about the employee’s performance corresponding
to these values on its screen. For example, to select Behavioral Alternative C under
the dimension French Fry Production, the subject would type the numeral 6, French
Fry Production in the list of dimensions, followed by the numeral 3 for Behavior C
in the list of alternatives. At this point, the computer would display the description
of the employee’s performance corresponding to Behavioral Alternative C on
French Fry Production (see Figure 7).

At the bottom of screen displaying the performance description, subjects
were asked whether they needed more information about the employee’s
performance. If they responded that they did, the program returned them to the
information selection screen with the opportunity to conduct another search. If they
responded that they were ready to rate the employee’s performance, the computer
program displayed a screen asking them to enter their ratings on a Likert-type scale.
While not a focal variable in the present study, these ratings were collected

primarily to maintain the face validity of this simulation.

33

 

Employee discards cooked fries held in the bagging station longer than
7 minutes

ENTER 1: IF YOU NEED MORE INFORMATION
2: IF YOU ARE READY TO MAKE THE FINAL RATINGS

 

 

Figure 7. Sample of Employee Performance Description.

Premieres

This study used two basic data collection methods--a self-report
questionnaire and a computer task which automatically recorded subjects’
responses.

Questioortaimompietion. Subjects reported individually for this study.
After entering the experimental room, they were seated at a desk on which an IBM-
compatible microcomputer was installed. Prior to beginning work on the computer
task, subjects were briefed concerning the general purposes of the study, and were
asked to sign a form indicating that they agreed to participate. (The consent form
appears in Appendix A.) Next, one-half of the subjects were randomly selected to
complete a questionnaire concerning their background as an employee in a fast food
restaurant, which was followed by working on the computer task (see Appendix B).

The remaining half of the subjects worked first on the computer task, and then

 

34
completed the questionnaire. This was done to control for any possible order or
cognitive priming effects associated with completing one of the tasks prior to the
other.

WW. Introduction and practice. At the
beginning of the computerized information search task, subjects were presented
with a message on the computer’s screen instructing them that they were about to
complete a simulation in which they were to play the role of a fast food restaurant
manager who must evaluate the performance of one of his or her employees. A
tutorial program, adapted from Schechtman (1987) assisted subjects in becoming
acquainted with the procedure for accessing information from the computer about
their subordinate’s performance (see Appendix C for the introductory messages and
tutorial on how to perform the computer task). Following this introduction, subjects
were given the opportunity to practice the search procedure by evaluating the
performance of an hypothetical college professor (see Appendix D for examples).

Four behavioral alternatives within each of four performance dimensions
were available for subjects to search about this professor. These dimensions
included Lecturing, Answering Questions, Advising, and Exam writing. All
dimensions were constructed so that they depicted average performance for a
college professor. Demographic characteristics of the professor were not provided.
Subjects were instructed that they could look for as much information about the
professor’s performance as they desired. They were also be told that they would
rate his or her performance after completing the information search portion of the
task. When subjects completed this practice portion of the computer task, the task

to evaluate the fast food employee began.

35

Introduction to the information search task. After familiarizing themselves
with the task of selecting information from the computer, subjects were
reintroduced to their task of searching for and evaluating the performance of a
subordinate fast food employee. At this time, the computer displayed a list of the
job dimensions among which subjects would search for performance information
about their employee (see Appendix E). Subjects also were informed that they
should carefully select the pieces of information they choose since they would only
be allowed to access one-half of all possible information about their subordinate’s
performance. Previous research with this task has shown that restricting the search
is necessary to prevent subjects from observing all possible information (M. L.
Doherty, personal communication, March, 1988). Once subjects read these
instructions, they were instructed to press the enter key on the computer’s keyboard
to begin the focal information search task.

Search task. In the computer search task, subjects were able to display
information about their hypothetical subordinate’s performance on six job
dimensions with seven behavioral alternatives within each dimension. Subjects
selected each piece of information they wanted displayed by entering one number
corresponding to the job dimension, and another relating to the behavioral
alternative. The order in which dimensions and behaviors were presented in each
list were randomized for each subject by the computer program.

Performance stimuli. The job dimensions and behaviors used in this study
were based on actual job descriptions and performance rating materials used by two
major fast food restaurant chains, and on the author’s interviews and on-sight

observations at two local fast-food restaurants. The descriptions of the hypothetical

36
employee’s behaviors were developed by selecting job dimensions and tasks which
were implemented in similar ways by both of the corporations for which job
descriptions were available. Six job dimensions, or "stations" were common across
the restaurants. These were: Dining Room Maintenance, Counter Service,
Sandwich Preparation, Outside Maintenance, French Fry Production, and Restroom
Maintenance. Next, the individual tasks performed within each dimension were
compared across restaurants. This examination revealed at least seven job tasks
within each dimension that were equivalent across the fast food chains used in this
study.

Once these 42 job tasks were identified (7 tasks within each of 6 job
dimensions), a behavioral statement was written to represent an hypothetical
employee’s performance on each task. So that there was no systematic bias in how
favorably any particular job dimension was depicted, behavioral statements were
written such that half were randomly selected to be worded favorably (i.e. effective
performance) and half to describe ineffective performance. Coincidentally, this
random selection resulted in three dimensions containing four positively-worded
descriptions and the other three dimensions containing four negatively-worded
descriptions. Descriptions were also worded in such a way that they were neutral
with respect to the sex of the hypothetical employee. These descriptions are
presented in Appendix F. Finally, the order in which the behavioral incidents were
entered into the initial list displayed by the computer was also determined randomly
within each dimension, and the computer program itself re-sorted the order of the

descriptions after each subject completed the search task. This system was designed

37
to avoid systematic effects due to the order in which the dimensions and behaviors
are presented.

All subjects were instructed to search for as much or as little information as
they felt necessary to make an accurate assessment of the hypothetical employee’s
performance within the constraints of the "one-half" search limit. Following the
completion of the search task, a hard copy of each subject’s search was created by
the printer connected to the computer. This information was not available to the
subject.

i ir m r

Self-rated performance. Subjects were presented with the same 42 tasks
which were used to develop the descriptions of the hypothetical employee’s
behaviors, and were instructed to rate their own performance on each of these tasks.
From these ratings, a score for a subject’s self-rating on each dimension was
computed using the mean of their ratings for each individual task within the
dimension. The self-evaluation instrument is presented in Appendix G.

Dimension impottanee ratings. Importance ratings were obtained for each
dimension by asking subjects to evaluate the importance or criticality of each of the
42 job behaviors mentioned above. An index of each dimension’s importance was
‘ calculated using the mean of the individual task importance ratings for
corresponding to each dimension. The instrument which was used to collect these
ratings is presented in Appendix H.

W. In order to capture different possible aspects of work
experience, three indices of a subject’s experience working the various job

dimensions to be rated were imbedded in the demographic questionnaire. The first

38
experience index was simply the subject’s tenure as a fast food restaurant employee.
’ A second index, corresponding to a subject’s experience performing each of the 7
job dimensions was obtained through self-reports of the amount of experience the
subject had had performing each dimension. This scale asked, "... please indicate
how much experience you have had performing the duties associated with each of
the stations listed." Responses on this item ranged from "no experience at all," to
"one of the most experienced employees in the store." Finally, an index of the
relative amount of experience a subject obtained across the job dimensions listed
was a report of the approximate percentage of time per week, on average, he or she
performed each job dimension when he or she terminated employment at the
restaurant (or for employees currently working in the restaurant, the percentage he
or she presently performed these tasks).

W. To identify any systematic differences in
search, rating, importance, and experience variables related to subject demographic
characteristics, questionnaire items asked for the name of the restaurant for which
the subject worked, if he or she was currently employed in that position, how long it
had been since he or she had terminated employment if no longer working for the
restaurant, as well as subjects’ ages and sex. In addition, the measure also included
an exercise in which subjects sorted the 10 tasks they performed best while a fast
food employee into the six job dimensions included in this study. This served to
facilitate subjects’ recall of their experience working in a fast food restaurant, and

was not analyzed as part of this study (see Appendix B).

39

W
Seatoh;_Amount. One index of a subject’s information search strategy was

the amount of information he or she searched about the hypothetical employee’s
performance within each job dimension. This variable was operationalized as the
number of behavioral incidents a subject displayed about each job dimension. This
information was obtained from the printed summary of each subject’s search
sequence.

WELLI- A second index of subjects’ information search strategy
involved asking them which job dimension they would select information about their
hypothetical employee’s performance from under the most severe information
search limits. Specifically, after subjects had completed the computerized search
task, they were asked, "If you could only view one piece of information about your
employee’s performance, from which dimension would you select it?" Subjects’

responses were recorded on a data summary sheet by the experimenter.

RESULTS

This section describes the analyses performed on the data collected in this
study. These analyses include an examination of the psychometric properties of the
self-rated performance and dimension importance scales, the effects of the control
variables Where Employed and Presentation Order on the primary research factors,
and finally the tests related to the proposed hypotheses.

W

Reliabilities for the self-rated performance and dimension importance scales
used in subsequent analyses are presented in Table 1. All reliabilities were
calculated using Cronbach’s (1951) coefficient alpha of internal consistency. Among
the scales, reliabilities for the self-rating scales ranged from an alpha of .76 for
Outside Maintenance to .95 for Restroom Maintenance. For the dimension
importance scales, reliabilities ranged from an alpha of .78 for Counter Service to
an alpha of .92 for Restroom Maintenance. These reliabilities indicate that all
scales had sufficient internal consistency for use in this research (alpha 3 .70;
Nunnally, 1978).

Tables 2 and 3 show the correlations among the self-rated performance
("Self-rating") and dimension importance ("Importance") scales, respectively. It
should he noted that the self-rated performance scales are moderately to highly

intercorrelated, as are the dimension importance scales. This suggests that the

40

41

Table 1.

Internal Consistency Reliabilities (Coefficient alpha) for Self-rating and Dimension

 

 

 

 

 

Importance Scales.
Measngrnent Scale
Self-rating Dimension Importance

19b Dimension Mean SD, alpha Mean 5.2, alpha
Restroom Maintenance 3.00 2.34 .95 3.71 1.12 .92
Dining Room Maintenance 1.58 1.70 .80 4.06 0.95 .85
Counter Service 5.49 1.79 .90 5.13 0.77 .78
Sandwich Preparation 4.65 2.33 .90 5.65 0.73 .82
Outside Maintenance 2.38 1.66 .76 3.50 1.11 .88

French Fry Production 4.74 1.96 .90 3.99 0.99 .85

 

Table 2.

42

Correlations among Self-rated Performance Scales.’

 

 

 

Sealepnnemion 1 2 3 4 5 6
1. Dining Room Maintenance --
2. Counter Service .49" --
3. Outside Maintenance .47" .14 --
4. Sandwich Preparation .23' .11 .32” --
5. Restroom Maintenance .64” .23' .59" .25" --
6. French Fry Production .50" .67" .30" .34' .39" ~-
‘ n = 101
° p < .05

"p < .01

Table 3.

43

Correlations among Dimension Importance Rating Scales.’

 

 

 

Seale Dimension 1 2 3 4 5 6
1. Dining Room Maintenance --
2. Counter Service .44" --
3. Outside Maintenance .53" .17 --
4. Sandwich Preparation .24' .13 .30" --
5. Restroom Maintenance .67" .20' .62" .26" --
6. French Fry Production .49" .68" .33" .39" .41" --
' n = 101
° p < .05

"p < .01

44

discriminant validity within each set of scales is low despite the fact that all scales
are internally consistent.
Control variables

A number of variables were inspected to see if they might cause spurious
relationships among the primary variables under study. Specifically, subjects’ age,
sex, how long it had been (in months) since they terminated their employment in the
restaurant, the restaurant in which they were employed ("Where Employed") and
the order in which the job dimensions were displayed on the computer’s information
selection screen ("Presentation Order") were coded and their relationships with the
principal research factors were examined. Presentation Order was represented
using six variables, one corresponding to each job dimension. These variables were
coded such that the variable listed first on the dimension selection screen for each
trial was coded "1," the second dimension listed was coded "2," etc. Where
Employed was coded using effects coding (Cohen & Cohen, 1983). Effects coding is
a procedure by which group membership, such as in which restaurant a person
worked, can be examined in a regression analysis. Two variables were required to
represent where subjects worked. Subjects who worked for Burger King were coded
"1" for the first variable and "0" for the second variable. For subjects with
McDonalds experience, the first variable was coded "0", and the second "1." Finally,
subjects who had worked for Wendy’s received "-1" codes for both variables. This
coding scheme allows the analysis of effects due to group membership in general,
rather than the comparison of results between groups.

The overall relationships among the primary research variables and the

variables representing the control factors were summarized using canonical

45
correlation analysis. The results of these analyses are summarized in Table 4.
These analyses revealed several significant relationships. First, age was significantly
related to the dimension which subjects reported they would select if they could only
search for one piece of information about their subordinate’s performance ("One
Piece"). In particular, older employees were more likely to select one piece of
information regarding Outside Maintenance than were younger workers. Next,
significant relationships were discovered between subjects’ sex (coded "1" for
females and "2" for males) and several research variables. Specifically, it was found
that females generally rated their own performance higher on Counter Service and
French Fry Production than did males, who rated their performance more favorably
on Outside Maintenance and Sandwich Preparation than did females. In addition,
females tended to: (a) report being relatively more experienced than other
employees, (b) spent a larger proportion of their time working, and (c) look for one
piece of information regarding Counter Service more frequently than did males.
Males, on the other hand: (a) reported being more experienced than other
employees, (b) spent a larger proportion of their time, and (c) tended to select one
piece of information regarding Outside Maintenance and Sandwich Preparation
more frequently than females.

The third control variable, the amount of time since subjects’ worked in a
fast-food restaurant ("Lapsed") was found to be negatively correlated with Relative
Experience and Tenure such that subjects who had worked longest in the restaurant
and rated themselves as most experienced had terminated their employment most

recently. Lapsed was also positively related to One Piece, with subjects who had not

 

46

 

 

 

Hoo. v a 2
Ho. v a :
mo. v a .
Hoa u m .
.me. om. mm. em..:me. :oe..:m¢. woman one .NH
an. He..:mm. mm. Hm. 6H. ma. oesoummm Hones: .Ha
mm..:mm. mm. no. :Hm. no. mo. ounces .oH
:om..:me. mm. mm. om..:ms. mm. emxuoz coflueum unmoumm .m
.He..:mm. em..:oe. :He..:~>. em. mocmfinmaxm m>wueaum .m
it tom. we. on. um. em. «a. mocmuuoaeH coflmcmeao .e
in mm. :44. om..:ee. ma. mocmeuouumm owumuuuamm .e
in em. NH. mm. mm. uueuo coflumucmmmum .m
it Ha. no. mo. emaoaaem mums: .e
it .em..:ao. acumen .m
in .ma. xmm .m
in mm< .H
h m m e m llml a . uouomm
..m.HOu.Um.m gunmmmmm HMEHQQ m 0.90Hm H00. MUM 0 m0

.v manna

47
been employed in a restaurant for the longest time being most likely to select
information about Outside Maintenance. Fourth, where subjects were employed
(Where Employed) was associated with Self-rated Performance and Relative
Experience. This suggests the possibility that subjects may have interpreted the
rating scales used for these dimensions differently depending on the restaurant for
which they worked. Finally, a significant relationship was found between
Presentation Order and Number Searched, with subjects searching for more
information within dimensions listed higher on the information selection screen
than for those towards the bottom of the list (see Figure 6 for the selection screen).
Therefore, because of their significant relationships with the primary research
variables, Age, Sex, Lapsed, Where Employed, and Presentation Order were
statistically controlled (partialled) in all subsequent analyses.

Table 5 presents the means, standard deviations, partial correlations, and
percent of variance partialled for all variables used in this study. The zero-order
correlation matrix is included in Appendix I. With respect to the number of pieces
of information searched, it should be noted that 65 percent of the subjects (n = 66)
searched for the full amount of data allowed (21 pieces of information). On the
average, subjects searched for 3.1 pieces of information per dimension (SD = 0.82),
with the mean number of pieces searched per dimension ranging from .17 to 3.5

pieces.

48

mN. «0.- ~N. co. co. o—. «m. me. On. vp. me. on. ~—. 00., on.~ mucoCOuc_ax ou_mu:o .op
~o.- .- mp.- we. m~.- «e. n~. oo. oo. no. Nn. oe. QN. o~.— 0¢.m aum>uom Leucaou .m—
NN. m—.- NF. ~o.- ow. —~. mm. on. on. op. p0. ow. me. on.. om.q oucaCOucmax loom acmcmo .qp

oucaacowcoa ounce-..om

 

 

 

me. o—.- «a. on.. up. we. ~o.- me. we. o—.- «o.- oc.- —F. oo.~P n¢.—p co_uu:u05a >uu cocoa; .m—
-- no.- am. pm.- n~. po.- cc. no. 5.. «9.- 9.. NP.- mp. ~m.e mo.n eucacuuc_e: acacumeu .N—
.- .9. so.- .o.. «p.- po.- Km. «o. mm.- 05.. ~—.- on. po.- oo.o~ co_uaceaeca cu_3ucam .pp
- on.- 0.. mo.- me. so.- on. m~.- oo- oo.- cc. mo.e om.~ eucecouc_o: ou_mu:o .op

- ne.- «o.- 09.- mn.- mp.- no. oo.- om. um. n—.on —~.~m oum>com coucaou .o

- on.- «o. c~.. ~c. c~.- pm. -.- mo. -¢.o 09.0 oucocoucmo: soon acmc_a .o
vogue: accuuom

- Pm. cc. nN. mm. m~. cc. no. ~m.— 0c.n composuOca >uu cocoa; .s

- Nm. ow. o—. —O. mm. ~«. 90., m~.~ oucacouc_e: eooLuuoa .o

- o~. o—.- o—. mm. so. oo.p «o.n comuocoaoca =u_IDCum .m

-- co. ~e. mm. no. mc.p on.p ouceceucma: oo_mu:o .o

-- w~. me. «o. Pm.— om.n ou_>cum coucaou .n

-- em. FF. oq.— o—.m oucmcouc_e: secs oc_c_o .~
oucu_eoaxo o>_ua.o¢

- op. we.o m~.—F ounce» .F

2 : 2 o o k a m .c m N _ 83$me dd cm... 238...:
.ce> a

 

keno—35m 373:3 :6chva 83%.; 5383— .9255 waos mecca—280 EEK

.m 035.

49

.—n

00000000 0002::

.m~
.eN
.nN
.-
.—~
.o~

00:0.0OQE. commces.o

 

.0.
.m—
.5—

 

0..—.500!

00. 00.- 00.- 0.. 00. 0.. 00. 00.- 0.. 00. 00. 00. 00. 0..- .0. 0.. 0.. 00.. 00.0 00..000000 .00 000000
00.. ...- 00. 00. .0.. 0.. 00.- 00.. 00. 0... 00.. .0. .0.- 0... 00. 0.. 00. 00.. 00.0 000000.0.0: 0000.000
00. 00. 00. 00. 0... 00. 00. 00. 0..- 0.. 00. 00. 0.. 0.. 0.. 0.. .0. 0.-0 00.0 00..0000000 00.00000
00.. ... 00.- 00. 0..- 00.- 0.. 00. 0..- 00. .... 00. 00. .0.- 00. 0.. 00. 00.. 00.0 000000.0.0: 00.0.00
00. 00.. ...- 00. 0... 0... 0... 00. .0.- 00. 00.- 00. 0.. 0.. 00. 00. 00. 00.0 .0-0 00.>000 00.0000
00.. 00.- 00.- 00. 00.- 00. 0... 00.- 00. ...- 0..- 0..- ...- 0..- 00. ... 0.. .0-0 00.0 000000.0.0: 0000 00.0.0
00. 00. 00. 00. 00. 00.. 00. 00.. 00. 00. 00. 0.. 00. .0. 0.. 0.. 00. 00.0 00.0 00..000000 .00 000000
00. 00. 00. 00. 0.. 00. 0.. 0..- 0.. 0.. 00. 00. 00. 00.- 0.. 00. 0.. 0... .0.0 000000.0.0: 0000.000
00. 00. 00. 00. 00. .0.- 0.. 0..- 0.. 00. 0.. 0.. 0.. 00.- 00. 0.. 00. 00.0 00.0 00..0000000 00.00000
00. 00. 00. 00. 0.. 00.- 00. 0..- 00. 0.. 00. 0.. 00. 00.- 00. ... 00. .... 00.0 000000.0.0: 00.0.00
00. 00. .0. 00. 00. 00.- 00.. 00.- ... 00. 00. 0.. 00. 00.- 00. 0.. 00. 00.0 0..0 00.0000 00.0000
00. 0.. 00. 00. 0.. 00. 00. 00.- 0.. 0.. 00. .0. 00. 0..- 0.. 00. ... 00.0 00.0 000000.0.0: 0000 00.0.0
00. 00. 00. 00. 0.. 00.- 00. 0.. 00.- 00. 00. 00. 00. 00. 0.. 00. 0.. 00.. 00.0 00...:0000 .00 000000
00. 00. 00. .0. 00. .0.- 00. 0..- 00. 0.. 00. 00. 00. .0. 00. 0.. 00. 00.0 00.0 000000.0.0: 0000.000
00. 00. 00. 0.. 00. 00. 0..- 00.- 0..- 00. 00. 00. 00. 00.. 0.. 00. 0.. 00.0 00.0 00..0000000 00.20000
A-u..couv eucaEco0cea uo.a.-.om
.0. 0. 0. 0. 0. .. 0. 0 .00 0 0 0 0 0 0 . 000.0...00 .0.0 0000
.0e> a

 

.0008. 0 0.00..-

50

ll
I-I

.0. v 0 .00.
00. v 0 .0.. u
00 0 .

Ll

C

 

00..0:0000 000 000000 .00
oucacouc.ox 2600000: .on
co_uo0aQ000 zu_3ucam .mn

oucaco.c.ox 00.0.30 .qm

ou.>0om 00.0309 .nn

 

 

00.. 00. 00. 00. 00.- 00.- 00.- .0.- 00. 00. 0.. 00.- 00.- ... .0. .0. .0. 0..0 .0.0

... .0. 00. 00.- 00.- 0..- 0..- 00. 0... 0..- 0..- 00.- 00.- 0.. 0..- 00. 0.. 0..0 00.0

.0.- 0..- 00.- 00. 00. 00. .0. 00.- 00.- 0.. 00.- 00. .0. 00.- 00.- 00.- 00. 00.0 0..0

00. 00. 00.- 0.. 00.- 00.- 0... 00.- 00.- 0.. 0.. ... 00. 00. 0.. 00. 00. 0..0 .0.0

00.. 0.. 00. 00.- .... .... .0.. 00. 00.. 00.- 00.- 0.. 00.- 00. 0.. 0.. 0.. 00.0 00.0

00. 00.- .0. 00.- 0.. 00. 0.. 0..- 00. 0..- 0.. 0..- .0. 00.- 00.- 0..- 00. 00.0 00.0 000000.0.00 0000 00.0.0 .00
smug-Cot; *0 0002a 0:0 .000 Buadom Emma—gun

0. 00, 0. 0. 0. .. 0. 0 0 0 0 0 0 0 0 . 000.0..000 .0.0 000:

«0:000:

.00) a

 

00.008. 0 0.000

S].

co_uu:uoga >L. gucogu ..m
mucmcauc_o: snagumoa .9n
co_uogoa¢ga cu_:ucam .o~
oucwcouc_o: on.nu:o .9~

oum>gom Loucaou .n~

vocugoom Lanes:
co_uu:uoLa >5“ scan-u .m~
oucocoucwu: EOOgunou .ew
comuogoabga somzucow .nN

ouccc0ucmox ou_ou:9 .-

00m>50m LOUCSOU .FN

oucougons_ co_mcas_o

vocacoucmo: Buchanan .9,

co_uogonoga gumzuccm .np
uuCgELOVLoa vouch-v.0m

 

 

- on. F9. on. 99.- 99.- F9. P9.- 09.- 99.- 09. N9. ~9.- n9. n9.
- m~.- M9.- mp.- 9n. m9.- N9. «9. <9. '9. 99. 99.- 59. «9.
- mp. ON. 99.- 59.- 59.- m9. «P.- P9. 09. 09. n9. M9.
- m9.- MN.- 99. 99.- 99.. 99.. m9. c9. MP. 99. 09.
-- 9—. <9.- 09.- PP.- .- PF. 9.. 99.- 99. «9.
-- op.- N9. «9. M9. N9. 99. o~.- or. «9. oocacmucmnx zoom newcmo .o~
-. mo. «0. ms. mm. 90. 9m. on. em.
-- F0. 90. F9. 09. ow. on. kw.
- No. mu. «0. 9n. -. MN.
- 90. m9. 9~. 9m. «N.
.- 90. on. mm. mm.
-- NN. um. ¢~. mucacuucmux soon uc_c_9 .9~
- ~¢. «e. comuuanoLa >Lu zucogu .op
- ON.
.9 on o~ 9w NN om mN «N nw - «N 9~ or 9' up

QLDmOOI

 

3.2.89 m 2.5

52

 

 

 

«a. .... no. me. oo.. ... no.- ~9. cc. NP. 0.. go. no.- we.. .9. co_uu:uo.a >5. gag... .Nn
qo.- op.. on. no.. 0.. ...- mo. ~F. ~o.- no. ho.- «p. no. .o.- so.. aucacoH:_ox soak...“ .om
~.. oc.- so. Po. o~.- -.- up. ~o. on. co. KN. n.. oo. no.- mo.- co_uagmauga =u_:ucom .mm
~o.. po. me. o... «o. o~. mp.- No.. .o.- no. No. me. «o.. mo. .9. oucaco.c_ax ou_.uao .em
...- no. oo. oo. ~F. ac. ~,.- op.- o~.- a..- m~.- ¢~.- cc. ma.- .9. uu_>..m ...caou .mn
«a. -. o~.- o,.. mo.- QN. mo.- mp. no. n.. Na. ~_. m..- Np. «o.- uucacch_oz goo“ mc_c_a .~m

smug-.02...— 50 000:. 060 so» guuodom CommCQ-mo
.n on o~ mm, - o~ m~ «N n~ - .N o~ a, a. s. «Lanna:

 

3.23 m 29.9

53

 

 

 

-- M9. 59.- m9.- 9—.- «9.- comuu3005a >5; gucugm .hn
-- 09.- 99. on.- 99. oucocouc_ax quagumua .om
-- 8.- K: 8.- 82239... 533:8 .mn
- n—.- «9.. oucwcouc_ox ov_mu:o .em
.- on.- ou_>gom goucsou .nn
- oucacoucmnx loom newcmo .~n
co_uaago.c_ .0 090.9 0:0 50* uo»uu.oa co_mca§_9
t. on m... en 3 [m 238..

 

9.389 m 03$

S4
Hypothesis-Que

Hypothesis One stated that a positive linear relationship should exist
between the favorability of one’s self-rated performance on a job dimension ("Self-
rating") and how important one perceived that dimension to be for successful overall
job performance ("Importance"). A canonical correlation analysis was conducted to
examine the relationship between Self-ratings and Importance at the global level
(across all job dimensions). The results of this analysis, presented in Table 6,
revealed a strong relationship between Self-ratings and Importance at the global
level, Em, (36,498) = 1.67, p = .01. Next, redundancy analysis was performed to
discover how much variance in Importance could be predicted from subjects’ Self-
ratings. Unlike squared canonical correlation coefficients, which represent the total
variance shared by two canonical sets, redundancy coefficients indicate what percent
of variance in one canonical set can be predicted by the other (as does 13} in
multiple regression). This procedure indicated that 24.2 percent of the variance in
Importance could be predicted from subjects’ Self-ratings, Em(6,80) = 4.25, p g .01;
confidence interval .10 - .39, p g .05. Finally, bivariate correlations were computed
between Self-ratings and Importance for each dimension. These analyses revealed
moderately strong bivariate correlations for four of the six job dimensions, with the
correlations on all six job dimensions being significant.

The hypothesized linearity of the relationship between Self-ratings and
Importance was examined by squaring each of the dimensional self-ratings and
adding these terms to the canonical analysis. If the relationship between these two
factors was curvilinear, then the addition of the squared Self-rating terms to the

predictor side of the canonical equation should have explained a significant amount

55

Table 6.

Canonical and Bivariate Correlations and Redundancy Index Between Self-ratings

 

 

 

and Task Importance.
Eature of Test
Global:
3.: Self-ratingjmponance .60
2
Bay ImportanaSelf-rating .24
Specific:
I Sell-rating reatroorn maintenanceJmportance restroom maintenance .36”.
L Self-rating dining room maintenanceJmportance dining room maintenance -44“.
L Sell-rating counter cerviceJrnportance counter service .33
I Sell-ratingaandwich, , " ' T ‘ aandwichpreparation 023"
L Sell-rating outside maintenanceJmportance outside maintenance .30 u
I Self-rating french try productionJrnportance trench try production 050

 

p<.05
° p<.01
' p<.001

56
of additional variance in the Self-ratings - Importance relationship. Results of this
procedure indicated only a 5.5 percent increase in variance explained, Em(1,99) =

1.20, n.s.; 1346,83) = 0.96, n.s.; confidence interval = .13 - .45’. Therefore, no

 

2 Because no straightforward test currently exists to test for significant
differences among variances (in this case redundancy indices; J. Hollenbeck, personal
communication, November 10, 1988), three separate tests were performed in an
attempt to converge on a conclusion. The first test implemented was the E... test,
described by Winer (1971). This test consisted of computing the ratio of the
variances obtained at each step (variance 2/variance 1), and then determining if this
value was greater than the critical value of the Em distribution for (2,99) degrees of
freedom.

The second test used was Cohen’s General E test for an Increment (Cohen &
Cohen, 1983). This consisted of solving the equation:

F (szut ‘ RzYA)/k3

...irtc

 

(1 ' R2YAB)/(n ’ kA ' kn ' 1)

where 11 corresponds to the number of subjects in the study, k, corresponds to the

number of predictors currently in the equation, 15,, to the number of predictors being

added to the equation, 133,, to the variance in X explained by predictors already in

the equation, and Em, corresponds to the variance accounted for all predictors after

the second set of predictors is added to the equation. The value of E resulting from

this equation was then compared to the critical value of the E distribution at (km n -
kA - k, - 1) degrees of freedom.

The final procedure used was to create 95% confidence intervals about the
greater (second) of the obtained variances and to assess if the lesser variance was
outside of this interval. (This procedure, recommended by Jacob Cohen in a class
at New York University, was communicated to the author by J. Hollenbeck,
November 10, 1988.) Confidence intervals were computed by approximating the
variance of B} using the formula:

var 3’ = 4[R2 (1 - R’)]2 / n

and then using this value to establish confidence intervals about the variance
obtained in the second step using the formula:

- g2 -1.96(var B’) s B’ s R’ + 1.96(var 3’)-

A significant difference was determined to exist between two steps if at least two
of these‘analyses was significant, and the third approached significance in the same
direction.

57
significant curvilinear component was discovered in the relationship between Self-
ratings and Importance, and Hypothesis One was fully supported.
W

Hypothesis Two stated that raters should search for more information about
another’s performance on job dimensions for which they evaluated their own
performance most favorably than on those for which they perceived their own
performance to be least favorable. Two canonical correlations were computed to
test this hypothesis: one between Self-ratings and Number Searched, and one
between Self-ratings and One Piece. As Table 7 demonstrates, neither of these
canonical correlations were significant: Em, (36,498) = 0.74, p = .86; and Em,
(36,498) = 0.99, p = .49, respectively. Examining the relationship between these
two concepts within each dimension revealed no significant bivariate correlations
among the 12 possible. These analyses demonstrate that Hypothesis Two was not
supported by the data.

Wm;

Hypothesis Three examined the relationships among self-rated performance,
dimension importance ratings, and information searched. Specifically, it was
hypothesized that the relationship between one’s self-rated performance on a job
dimension and his or her subsequent search for information about another’s
performance on that dimension is completely mediated by the dimension’s
perceived importance for successful job performance. Complete mediation suggests
that the relationship between an antecedent or "exogenous" variable (Self-rating)

and a subsequent, "endogenous" variable (information search) is completely

58

Table 7.

Canonical and Bivariate Correlations Between Self-ratings and Information Search.

 

 

Nature of Test
D’uitt'n «.11..- ' =Nmn‘ I 0' ' l." .1. it ' -01,
Global:
E: Self-ratingfiunber Searched .4 1
Specific:
I Self-rating restroom maintenanceN umber Searched restroom maintenance ”~07
I Self-rating dining rm. maintenanceNumber Searched dining rm. maintenance '-05
I Self-rating counter aervice.N umber Searched counter service '.04
I Self-rating sandwich preparationNun'ber Searched sandwich preparation '03
I Self-rating outside maintenance.N umber Searched outside maintenanc '-03
I Self-Intin‘nenehhyproductionfiumberSearchedhenehflyproduabn '-02
.""'!VI'1"I)110‘ 'H'ni " ‘tJro ‘011"ofor aiu
Global:
E. ...-...... .... .44
Specific:
I Self-rating restroom mintenanceOne Piece restroom maintenance -00
I Self-rating dining rm. maintenance.0ne Piece dining rm. maintenance ~07
I Self-ruin; counter service.0ne Piece counter service . 13
I Selfonting sandwich preparationflne Piece sandwich preparation “.05
I Self-rating outside maintenance.0ne Piece outside maintenance 006
I Self-rating french fry production.0ne Piece french try produaion '-05

 

p<.05
p<.01
p<.001

59
transmitted or conveyed through a third, "mediating" variable which intervenes
between them (Importance).

Assessing if the relationship between two factors is mediated by a third
variable is a two-step process (cf. James & Brett, 1984). First, the significance of the
relationships between the exogenous, endogenous, and mediating variables must be
established. A significant correlation between the exogenous and endogenous
variable indicates that there is some type of relationship between these two
variables, while a significant relationship between each of these variables and the
hypothesized mediator provides a channel through which this relationship might be
transmitted. If each of the relationships between these three factors is statistically
significant, then the second step of the analysis is to partial variance associated with
the mediating variable from the relationship between the exogenous and
endogenous variables. This is typically accomplished using multiple regression to
regress the mediation variable on the endogenous variable, adding the exogenous
variable to the regression equation, and analyzing the statistical significance of the
change in total variance explained between the two equations (EMF). Complete
mediation exists if adding the exogenous variable does not significantly add to the
amount of variance explained. Understanding that Hypothesis Two revealed no
significant relationship between Self-ratings and information search, the following
analyses were conducted primarily to examine the relationship between information
search and Importance.

Figure 8 presents the canonical correlations among Self-ratings, Importance,
and each of the information search variables (see Figure 8). As revealed by

Hypotheses One and Two, the relationship between Self-ratings and Importance is

60

Figure 8.

Canonical Correlations Among Self-ratings, Importance, and Information Search.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.41 (na)

Salt-ratlng ’ 82:21::
.60 (.01) .39 (no)
Importance
.44 (no)

Self-ratlna C One Piece
.60 (.01) .48 (.05)

 

Importance

 

 

61
significant, Em, (36,498) = 1.67, p = .01; while the relationship between Self-ratings
and information search is not. The present analyses demonstrate that while the
relationship between Importance and Number Searched was not significant, Em,
(36,498) = 0.73, p = .88; the relationship between Importance and One Piece was
significant, Em. (36,498) = 1.54, p = .03. However, since the relationships among
all three variables under consideration were not significant (Self-ratings,
Importance, and information search), Importance cannot be considered a mediator
between Self-ratings and information search.

The nature of the relationships among Self-ratings, Importance, and
information search were also examined to see if the hypothesized mediation
relationship was dimension specific. These analyses revealed that, since the
correlations among all three variables were not significant within any of the
dimensions, no mediated relationships existed at the dimensional level. These
results indicate unequivocally that Importance did not mediate Self-ratings and
information search in this study.

W

Hypothesis Four asserted that the relationship between experience
performing the tasks associated with a job dimension and one’s search for
information about others’ performance on that dimension is mediated by the
dimension’s perceived importance for successful job performance. This hypothesis
was tested using three separate indices of experience. The first experience index
consisted of subjects’ ratings of how much experience they had performing a
particular station’s duties compared to other employees in their restaurant

("Relative Experience"). The second measure of experience was subjects’ estimate

62
of the percentage of time per week, on average, they were assigned to each job
station when they last worked at the restaurant ("Percent Worked"). The final index
of experience was simply the length of time subjects had worked as a fast food
restaurant employee ("Tenure").

To test the mediation hypothesis at the global level, it was first necessary to
establish whether significant relationships existed among experience, dimension
importance, and information search. Figure 9 displays the canonical correlations
among each index of experience, Importance, and each measure of information
search. The data summarized in this figure indicate that all three relationships are
not significant in any of the six sets of variables. Strictly applying p = .05 as the
significance level, it is concluded that Importance does not serve as a mediator in
any of these combinations of variables.

However, the canonical correlation between Relative Experience and One
Piece is marginally significant, with p = .07. To investigate if the mediation
hypothesis can be supported at any level, the possibility that this relationship might
be mediated by Importance was tested using redundancy analysis. The results of
this examination, shown in Table 8, indicated that the Relative Experience - One
Piece relationship was no longer significant with the effects of Importance
partialled, E... (1,99) = 2.01, p < .05; _E,,,,(6,77) = 1.61, n.s.; confidence interval .07 -
.33, us. This suggests that using the relaxed significance, Importance does mediate
the relationship between Relative Experience and One Piece. Therefore,
Hypothesis Four is conditionally supported at the global level.

Hypothesis Four was also explored by looking at the relationships among the

experience, importance, and information search variables within each dimension.

Figure 9.

Canonical Correlations Among Experience, Importance, and Information Search.

 

 

 

 

 

 

 

 

 

.56 (.07)

 

 

 

 

 

 

Experlence

 

 

 

 

 

 

 

 

 

.30 (no)
Wire A w
Elnerleeee ' son-om
.40 (.06) .39 (as)
W
.08 (no)
Percent _ Nunbet
w ' Searched
.54 (.06) .39 (m)
(W
.30 (no)
flatter

 

 

 

 

 

 

 

OnePlece

 

 

 

 

 

 

 

 

 

 

 

 

 

ti

 

.40 (no)
Peroant
W
.54 (.06)
heart-we
.37 (no)

 

 

 

 

 

 

 

“PI.“

.48 (.05)

 

 

 

 

 

 

 

 

 

 

Table 8.

Relationship Between Relative Experience and One Piece Partialling Importance.

 

2
BO] One Piece.Relative Experience .10
R 2 20
_dy One PieceJmportance Relative Experience '

Variance Relative Experience - > One Piece - Rafa... M. 1m...” ”...... Wm... -

2
Edy One Piccellelative Experience

.10‘

 

‘_m(1,99) = 2.01, p _<_ .05; Em,(6,77) = 1.61, p_ Z .05; confid. int. = .07 - .33, n.s.

65

The partial correlations presented in Table 5 report that all three relationship
between experience, information search, and Importance were significant only when
considering Percent Worked, Importance, and One Piece for Counter Service.
Partialling Importance from the relationship between Percent Worked and One
Piece revealed a correlation between Percent Worked and One Piece that was still
significant, “(136) = 4.20, p = .04. This indicated that Importance did. not
mediate the relationship between Percent Worked and One Piece, signifying that
Hypothesis Four is not supported at the dimensional level.
W

Hypothesis Five suggests that raters’ experience in performing the tasks
associated with a specific job dimension will affect their perceptions of that
dimension’s importance directly, as well as indirectly through their evaluations of
their own performance. As with Hypotheses Three and Four, the relationships
among all three factors must be significant for the hypothesis to be supported.
Canonical correlations were computed among Importance, Self-ratings, and each of
the three experience indices, and the results of these computations are presented in
Figure 10. These figures show that all three relationships were significant when
experience was represented by Relative Experience, as well as by Percent Worked.

Relative Experienee. Since all relationships among Relative Experience,
Self-ratings, and Importance were significant, the next step was to decompose
Relative Experience into its direct and indirect effects on Importance (cf. Cohen &
Cohen, 1983). This procedure involves first partialling the variance associated with
Self-ratings from the Relative Experience - Importance relationship, and likewise

the effects of Relative Experience from the relationship between Importance and

66
Figure 10.

Canonical Correlations Among Experience, Self-ratings, and Importance.

.48 (.05) .54 (.05)-

 

 

 

 

mm
"mu" - Importance Importance

Experience WWW

 

U

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.73 (.01) .60 (.01)
Self-refine
.22 (no)
Tents'e ; Importance
.65 (.01) .80 (.01)

Self-refine

 

67
Self-ratings. Since these relationships were analyzed at the global level, partialling
was performed by computing canonical correlations in a hierarchical manner, as in
multiple regression, and then examining the percent of variance explained in the
criterion using redundancy analysis. For example, to partial the effects of Self-
ratings from the relationship between Percent Worked and Importance, a canonical
correlation was computed between the set of variables representing Importance and
the set representing the Self-ratings. The redundancy index represented the percent
of variance that Percent Worked predicted in Importance. Next, the set of Relative
Experience variables were added to the Self-ratings set on the "predictor" side of the
equation, and the redundancy index represented the variance these two concepts
predicted in Importance. The change in variance explained between the first and
second equations represented the variance in Importance directly associated with
Relative Experience. This change in variance was examined using the methods
suggested by Hollenbeck in Hypothesis One to determine if Relative Experience
predicted a significant amount of variance in Importance (J. Hollenbeck, personal
communication, November 10, 1988). The results of these analyses, presented in
Table 9, indicated that, when Self-rating was partialled from the Relative
Experience - Importance relationship, this relationship was no longer significant:
Em(1,99) = 1.38, p g .05; En,(6,77) = 1.76, n.s.; confidence interval .15 - .51, n.s.
This suggests that Self-ratings completely mediated the relationship between
Relative Experience and Importance. Therefore, no test of the indirect relationship
between Relative Experience and Importance was conducted on these data, and
Hypothesis Five was not supported globally using Relative Experience as the index

of experience.

68

Table 9.

Relationships Among Relative Experience, Self-ratings, and Importance.

 

Canonicalmmm

KW... ......“ = .482
Rama...“ = .596
3...... Wm“... = .844
Redundancy Indiees

B.’............-.... 5...... = .139
BJ-wmm..., = .239
B...’-...........s.-....., ...... 5...... = .454

Hierarehieel Analysis

- Direct relationship between Relative Experience and Importance

 

partialling Self-rating
Variance REL - > IMP = Rq’,mm,s,mmmm Wm, - _Rq’mponmnswflm
= .331 - .239
= .092
Em = 1.38' Em(6,77) = 1.76 Confid. Interval R,2 = .15 -.51
° p < .05

"p < .01

69

W. The direct and indirect effects of experience on
Importance using Percent Worked as the index of work experience were examined
using the same procedures as for Relative Experience. These analyses, detailed in
Table 10, revealed that the amount of variance Percent Worked predicted in
Importance was not significant when variance associated with Self-ratings was
partialled from the relationship: Em(1,99) = 1.43, p < .05; Em,(6,77) = 2.04, n.s.;
confidence interval 16 - 52, n.s. Consequently, Self-ratings also completely mediated
the relationship between Percent Worked and Importance, making it unnecessary to
evaluate the hypothesized indirect relationship between these variables. Therefore,
Hypothesis Five was also not supported at the global level with Percent Worked as
the index of experience.

Dimensienel Analyses. Hypothesis Five was also tested at the dimensional
level. In examining the bivariate correlations among the three experience measures,
Self—ratings, and Importance within each of the six job dimensions, all three
correlations were significant for seven sets of variables: Dining Room Maintenance,
Outside Maintenance, Restroom Maintenance, and French Fry Production using
Relative Experience as the experience index; and Counter Service, Outside
Maintenance and French Fry Production using Percent Worked as the index of
experience. Since all three variables were significantly correlated in each of these
set, the next step was to partial Self-rating from the relationship between
Importance and experience. In all but one case, this step resulted in non-significant
relationships between experience and Importance (see Figure 11), indicating that

the relationship between experience and Importance was mediated by Self-rating in

70

Table 10.

Relationships Among Percent Worked, Self-ratings, and Importance.

 

Qanom’caLleations

BcIWercent Worked = .536

BelmonanceSett-rating = .596

3...... ...-......“ = .731

MW

Iii-mm... ...... ...... = .122

B.’-........ ...-...... = .239

fi-f-mm ...-...... ...... w... = .344

Hiererehieel Analysis

- Direct relationship between Percent Worked and Importance
partialling Self-Rating

. _ 2 2
varlance PER ‘ > IMP - Bay lmportanoe..Selt-rating Percent Worked ' Bay Importance. Sell-rating

= (.344 - .239)
= .105
13... = 143‘ 546,77) = 2.04 Confid. Interval R: = .16 - .52

 

'p< .05
"p< .01

71
Figure 11.

Dimensional Relationships Among Experience, Self-Ratings and Importance.

Outside Management

.25 (.01) .23 (.06)

 

 

 

 

um mm
4, Inmcrtanoe ‘ Imortaroe
Experience Worked

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 
    

.22 (.06) .30 (.01)

 

 

 

 

 

 

 

 

 

Self-rating
AR '-.01tm) An '- .02 (no)
French Fry Production
.33 (.01) .29 (.01)
new» _ Percent
Experience ' ”no“ Wound ' Importance

 

 

 

 

 

 

 

 

 

 

 

 
   
 

.28 (.01) .50 (.01)

 

 

 

AR ’- .00 (ne) AR '- .02 (no)

Figure 11 (cont’d.)

72

Restroom Maintenance

 

Dining Room Maintenance

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

26 (.01) .19 (.05)
Robtlve Relatlvo
m I
Experlonoe lull Hm”. mportanoe
.73 (.01) 30 (~01) .oo (.01) .44 (.01)
Self-mm SCH-ruby
AF) 1- .00 (no) A R" .01 (no)

Counter Service

 

 

 

 

 

 

 

 

 

 

-.2O (.06)
Percent _ I
Worked moortanoe
.42 (.01) .33 (.01)
Self-rating

 

AR '- .13 (.01)

 

73
all three sets. Therefore, no test of the indirect effects of experience through Self-
rating was made for these sets.

Looking at the relationship between Percent Worked and Importance with
Self-rating partialled for Counter Service, however, revealed that Self-rating did not
completely mediate the relationship between Percent Worked and Importance.
Therefore, the direct and indirect effects of Percent Worked on Importance were
examined. The direct effect of PerCent Worked on Importance was found by using
multiple regression to partial Self-rating from the Percent Worked - Importance
relationship. This revealed that Percent Worked directly accounted for 13.5 percent
of the variance in Importance, and therefore a direct effect of .367.

To determine the indirect effect of Percent Worked on Importance, it was
also necessary to obtain the direct effect of Self-rating on Importance. Partialling
Percent Worked from the Self-rating - Importance relationship using multiple
regression revealed a direct effect of .438 between Self-rating and Importance. The
indirect effect of Percent Worked on Importance was then calculated by multiplying
the direct effect of Percent Worked on Self-rating by the direct effect of Self-rating
on Importance. This resulted in a significant indirect effect of .184, Ew(1,99) =
1.25, p < .05; 541,85) = 3.48, p < .01; confidence interval .05 - .28. However,
despite finding that the hypothesized relationships existed in this case, a complete
mediation model better describes the general relationships among experience, Self-
ratings and Importance. Therefore, only minimal support for Hypothesis Five was

found at the dimensional level.

74

may:

Analyses using Number Searched and One Piece as the search indices
indicated that information search was not related to raters’ self-concepts. It is
possible, however, that another aspect of information search, the dimension from
which raters first chose to display information about their subordinate’s
performance, might be related to raters’ self-concepts. This possibility was tested by
computing canonical correlations between a set of dummy-coded variables
representing the dimension raters first displayed and the primary exogenous
variables of interest in this study (Self-rating, experience, Importance). Results of
these analyses revealed that the relationship between the dimension first selected
and Self-rating was marginally significant, 3, = .50, p = .08. No other relationships

with the dimension first selected were significant.

DISCUSSION

This study investigated the relationship between raters’ self-concepts and
their search for information about others’ performance on a job which they
themselves have performed. Representing those parts of their self-concepts which
relate to job performance, raters were asked to judge their own performance along a
number of behavioral dimensions, report how important they felt the tasks
comprising those dimensions were to successful job performance, and to indicate
how much experience they had performing the tasks under investigation. It was
hypothesized that there would be significant relationships among these various
aspects of raters’ self-concepts, and that certain of these features would influence
what information raters’ selected to view about an hypothetical other’s performance.
Specifically, it was hypothesized that raters’ ratings of their own performance on a
given job dimension would be significantly related to both how important they felt
that dimension was to successful job performance and to how much infOrmation
they would search for about another’s performance on that dimension. It was
further hypothesized that the perceived importance of a job dimension for successful
overall performance would mediate the relationship between raters’ information
search and both their self-ratings and their work experience. Finally, it was

hypothesized that raters’ own reported performance level would partially mediate

75

76
the relationship between their experience and how important they felt a job
dimension was for successful overall performance.
Writs

Hypothesis One predicted that raters’ ratings of their own performance on
the tasks associated with a particular job dimension would be significantly related to
their ratings of that dimension’s importance. This relationship was strongly
supported in these data both at the global and dimension-specific level. Self-ratings
predicted 24% of the variance in importance ratings at the global level, and up to
25% within any one dimension. This finding is congruent with previous research in
the social cognition literature regarding self-image bias, which suggests that the
more desirable one’s self-ratings on a dimension, the more central or important that
dimension is perceived to be (e.g. Lewicki, 1983).

The second hypothesis investigated the proposition that raters would search
for more information about their subordinate’s performance within dimensions on
which they rated their own performance most favorably. This hypothesis was not
supported, whether raters were allowed to search for several pieces of information
about their employee’s performance or only selected one piece of information.
However, the post-hoc analysis revealing a marginally significant relationship
between raters’ self-rated performance and the dimension which they selected to
view first on the search task suggests that one’s self-concept may, in fact, influence
attention to behavioral information. If this is indeed the case it would be congruent
with the self-image bias perspective of person perception, which implies that such an

effect should exist.

77

The third and fourth hypotheses each predicted that a job dimension’s
perceived importance would mediate the relationship between one’s search for
information about another’s performance and a central feature of the rater’s self-
concept--either the rater’s own performance (Hypothesis Three), or his or her
experience on the dimensions being rated (Hypothesis Four). Neither of these
hypotheses were supported in this study using conventional standards of
significance, primarily because no significant relationship was found between self-
rated performance and information search in these data. However, there was
support for the mediated relationship between raters experience on a dimension
relative to other employees in their restaurants and which dimension they selected
for one piece of information using 2 g .10. One possible source contributing for the
lack of additional significant findings related to information search is the search
methodology used in this study. More will be said about this issue later in this
discussion.

Finally, Hypothesis Five sought to determine if a direct relationship existed
between one’s experience working a job station and its perceived importance to
overall job success, and if one’s evaluation of his or her own performance partially
affected this relationship. It was discovered that the relationship between one’s
perceptions of a dimension’s importance and the percentage of time spent working a
station, as well as one’s level of experience compared to one’s peers, was completely
mediated by one’s self-evaluated performance. Since the hypothesized relationships
among these variables indicated only partial mediation, Hypothesis Five was not

supported by these data.

78

Overall, this study provides partial support for the application of both the
self-image bias and expertise perspectives of perception to the realm of perceptions
about behavioral information. The self-image bias perspective correctly predicted a
significant relationship between ratings of one’s own performance and ratings of
what was considered important to overall job success. The expertise perspective, on
the other hand, accurately anticipated that the greater one’s experience performing
a task dimension relative to other dimensions, the greater will be that dimension’s
perceived importance.

imi i n f tud

The primary focus of the present research was that of predicting rater’s
search for information about their subordinate’s performance. Unfortunately, little
of statistical significance was found regarding this variable. There are several
possible reasons for the lack of significant relationships. One conclusion may be
that no relationship exists between raters’ self-concepts and what they attend in
others with regard to behavioral information. Both the self-image bias (Lewicki,
1983; 1984) and expertise perspectives (e.g. Fong & Markus, 1982) described earlier
present evidence that people will attend to trait information about others which
corresponds with their own self-concepts. It may be, however, that this effect does
not generalize to one’s perceptions of behaviors. Behaviors are discrete acts,
occurring within a specified time and place. Trait descriptions, on the other hand,
depict general behavioral tendencies across various contexts and are therefore
better elaborated in memory. Because more information is represented by a trait

description, it is more likely that environmental stimuli will match one’s accessible

trait schemas and will therefore be attended to more readily. Memories of one’s

79
own behaviors, on the other hand, are less likely to correspond to other incidents in
one’s environment, and therefore are less likely to affect information acquisition.
These differences between traits and behaviors may explain the lack of any
statistically significant relationships between one’s self-concept and information
acquisition in this study.

A second possible explanation for the lack of significant relationships with
information search is that the method used to collect these data made it difficult to
detect systematic relationships between subjects’ self-concepts and the job
dimensions they chose. This problem appears to be centered around using the
version of the computerized information display board employed in this study.
Specifically, it appears that subjects selected information about their hypothetical
employee’s performance with the purpose of displaying an equal amount of
information about each job dimension. As reported earlier, sixty-six percent of the
subjects in this sample searched for the full amount of information allowed (21
pieces), searching for an average of 3.09 pieces of information per dimension (SD.
= .82). The average number of pieces of information accessed per dimension,
however, ranged from 2.34 to 4.21, both of which are within two standard deviations
of the mean. This data suggests that there was insufficient variance in the number
of pieces of information searched across dimensions to detect systematic
relationships between information search and the self-concept variables.

It is also questionable whether working as an employee in a fast food
restaurant constituted a significant portion of these subjects’ self-concepts when this
research was conducted. First, working in a fast food restaurant may be such a

simple job that minimal experience and effort are required. This might make it less

80
likely that fast food employees will significantly incorporate their jobs into their self-
concepts. Secondly, though subjects had worked in a restaurant for an average of
11.25 months, only 11% of the sample was currently employed in a fast food
restaurant, and it had been an average of 17.5 months since the rest of this sample
had terminated their employment. Therefore, even though subjects may have
incorporated their jobs as fast food employees into their self-concepts while
employed, their identification with that job may have dissipated over time. Partial
support for this view may be found in the relationship between the length of time
since subjects’ worked in a fast food restaurant and their ratings of their Relative
Experience. This relationship indicated that subjects rated the amount of
experience they had relative to other employees lower the longer it had been since
they had worked in a fast food restaurant. At very least, the task importance ratings
and the ratings of one’s own performance would have been more salient in such a
sample.

Several less-crucial limitations of this study exist as well. First, more
significant relationships may have been found with the dimension subjects selected
for one piece of information about their subordinate’s performance ("One Piece")
had there been more variance within each dimension. Except for Counter Service
and Sandwich Preparation, the standard deviations for the One Piece variables were
less than .20, with these variables ranging from 0.0 to 1.0. Another possible
limitation in this study involves the dimension importance ratings. In asking subjects
to rate how important each task was to successful job performance, questions were
worded such that subjects could have responded either how important they (the

subjects) thought the various tasks were, or how important the W they

81

worked for considered these tasks. A case in point is that, while a significant
relationship was found between Importance and One Piece (R, = .45, p = .05), the
correlation between One Piece and subjects ratings of the importance of Counter
Service, the dimension selected for one piece of information by 73% of the subjects,
was -.25, p = .01. Third, the population from which participants in this research
were drawn imposes another set of possible limitations on this research. All
subjects were given course credit to encourage participation in this study. Since
simply "participating" was necessary to receive this credit, subjects may have simply
"gone through the motions" rather than putting forth their best effort when
completing the required tasks. This may explain why 7 percent of the subjects (_n =
7) searched for less than 6 pieces of information total about their subordinate’s
performance--less than an average of one piece per dimension.
I 1i i f h rr n

The ramifications of the current study are limited regarding the process of
attending to and selecting information about the performance of others. The
existence of a marginally significant relationship between self-rated performance
and which dimension subjects selected for their first piece of information provides
limited evidence of the self-concept’s relationship to information search, and
therefore its implications for research and practice are currently minimal. On the
other hand, the relationship discovered between one’s perceptions of what job
dimensions are important and his or her own performance may prove important for
future research in which subjects are required to either rate their own performance,
or to rate the importance of job tasks which they perform. As demonstrated in the

current study, self-rated performance and perceived dimension importance are

82

highly related to a person’s experience performing the job being related, as well as
to each other. Therefore, any ratings of a task’s importance may be biased by how
well the person providing the rating performs on that task. Such a relationship
should be kept in mind when, for example, interviewing subject matter experts for a
job analysis. It is possible that the importance of each job duty will vary based on
how well each expert performs that duty. This highlights the importance of
collecting data from a number of experts before a final determination of a task’s
importance is made.
D' i f r r r r h

This study provides an initial exploration of how one individual difference
variable, a rater’s self-concept, might affect how he or she perceives the
performance rating context, and specifically how one’s self-concept might influence
what is perceived about other’s performance. Future research could continue this
line of inquiry in a number of ways. First, the nature of the relationship between
one’s self-concept and attention to others’ performance needs to be better
understood. The data collected in this study, especially the relationship between
Importance and One Piece and the marginally significant relationship observed
between Self-ratings and the dimension from which subjects’ selected their first
piece of information, provide initial evidence that the self-concept is associated with
attention to performance information about others. Further evidence is necessary,
however, to corroborate these findings and to explore other ways in which the self-
concept might affect behaviorally-based person perception.

Secondly, future research should attempt to establish a causal, rather than

simply a correlational, relationship between the self-concept and attention to

83
performance information. Though an expensive proposition, such research could
involve hiring subjects to perform a job over an extended time period. This would
make it possible to collect measures assessing how job-related self-concepts change
over time, are (affected by positive or negative feedback and alterations in task
difficulty or criticality to overall job performance, and how these changes affect
one’s information search process.

Another area in which subsequent research could improve upon the current
study relates to the methodology used. First, subsequent studies should focus on
developing alternative methods for measuring rater’s information search process.
While the computerized information display board is convenient in terms of data
storage and retrieval, it presents subjects with a highly artificial situation. Using this
system, raters read about, rather than actually observe, vignettes of ratee
performance. Actually observing real people perform the tasks to be rated would
allow raters to draw their own conclusions about ratees’ performance levels. This
may also make it more difficult for raters to decide if they have enough appropriate
information about their subordinate’s performance to make a judgment. If this
method does increase the amount of information needed to make a rating, self-
concept related effects may be more prominent in an expanded search for more
information. This video approach to information display might involve using a
computerized interactive videodisc system in which subjects would select which
behavior to view using the computer’s keyboard and then watch the corresponding
behavior on an attached video monitor.

If future research continues to use the current application of the information

display board, however, several changes should be made. First, the computer

84

program which controls this simulation should be modified to present subjects with
a list of dimensions and alternatives which are randomly ordered each time a piece
of information is displayed. This would help reduce the effects of presentation
listing order on how subjects choose to display information. Secondly, some
measure should be obtained to evaluate subjects’ use or "fear" of computers.
Currently, the effect of subjects’ affective reactions to working on a computer are
not regularly assessed. Therefore, it is impossible to judge whether interest or
experience with computers is affecting subjects’ information search processes.
Finally, subsequent research should not limit the amount of information subjects
can access about their employee’s performance. Not only does a search limit restrict
potential variance in the number of pieces of information subjects can search
overall, it also appears to encourage subjects to access the same number of items
within each dimension, also restricting variance. Since recent research has found an
increase in variance when subjects are presented with a very large number of
alternatives and dimensions (J. K Ford and S. Kozlowski, personal communication,
March 30, 1989), future studies should attempt to "overload" subjects with
information, thus enhancing the possibility of variance in search across dimensions.

A final area in which future work-related self-concept research could
improve relates to the selection of a sample. As mentioned earlier, the lack of
significant findings associated with information search in this study may be due in
part to the length of time since these subjects worked in a fast food restaurant. The
association between self-rated performance and the amount of time since working in
a fast food restaurant, for example, suggests that if subjects ever identified strongly

with their roles as fast food employees, then the intensity of this identification has

85

dissipated with time. This effect could be avoided in the future by recruiting
subjects who are incumbents in the jobs being evaluated. Secondly, the type of job
examined in this study may too simple to produce the expected results. That is,
casual observation as a customer in one of these restaurants may be sufficient
exposure to: (a) develop a working knowledge of what is important for job
performance, (b) assess how much experience one already possesses performing
tasks similar to those necessary for working in the restaurant, and (c) identify
approximately how well one can perform the tasks that employees do. Therefore,
future research on work-relevant self-concept issues would benefit from soliciting
subjects who are incumbents in meaningful, complex jobs.
Conclusion

The current study was an attempt to demonstrate the effects that a rater’s
self-concept has on the selection of information about another’s performance in a
work context. Though this effort was generally unsuccessful, some evidence was
found which suggests that one’s self-concept does affect attention to certain aspects
of performance-related behavior. Specifically, a significant relationship was found
between a rater’s evaluation of his or her own performance and what he or she
considered important for successful job performance. Significant relationships were
also found among a rater’s experience performing the tasks related to a particular
job dimension, his or her self-evaluated performance on those tasks, and how
important he or she perceives those tasks to be. Finally, limited evidence was
found linking information search to a rater’s self-concept. The absence of stronger
evidence supporting a relationship between one’s self-concept and information

search is hypothesized to be a result of inadequate methods used to assess the

86
information search process and the use of a less than optimal subject sample.
Additional experimental research which further explores the relationship between
the rater’s self-concept and his or her information search and which improves on the
sampling and data collection methods employed in this study is suggested. Though
the hypothesized relationships between self-concept and information search were
not found, this study provides a foundation upon which later examinations of

individual differences in rater attentional processes may build.

LIST OF REFERENCES

LIST OF REFERENCES

Bargh, J. A. (1982). Attention and automaticity in the processing of self-relevant

information. learnal ef Persenality and Seejal Esyehelegy, 43, 425-436.

Bargh, J. A., & Pietromonaco, P. (1982). Automatic information processing and
social perception: The influence of trait information presented outside of

conscious awareness on impression formation. Jearnal ef Persenality and
W. 4.3. 437-449.

Bargh, J. A., & Thein, R. D. (1985). Individual construct accessibility, person
memory, and the recall-judgment link: The case of information overload.

leprnal ef Persenality and Seeial Psyehelegy, 19, 1129-1146.
Bartlett, F. C. (1932). Remembering. London: Cambridge University Press.

Borman, W. C. (1978). Exploring upper limits of reliability and validity in job
performance ratings. Jearnal of Applied Psychology, 63, 135-144.

Bruner, J. S. (1951). Personality dynamics and the process of perceiving. In R. R.

Blake & G. V. Ramsey (Eds.), Pereeptien; An approaeh ta perspnality. New
York: Ronald Press.

Bruner, J. S. (1958). Social psychology and perception. In E. E. Maccoby, T. M.
Newcomb, & E. L. Hartley (Eds.), R in in i 1 ch 1 . New York:
Holt, Reinhart, & Winston.

Bruner, J. S. (1983). In seareh ef mind. New York: Harper & Row.

Cantor, N ., & Kihlstrom, J. (1986). Persenality and soeial intelligence. Englewood
Cliffs, NJ: Prentice-Hall.

Carver, C. S., & Scheier, M. F. (1981). Attentien and self-regulatieg; A eentrol
theoty appreaeh te haman behavier. New York: Springer-Verlag.

Cohen, J.&Cohen, P. (1983). A 1i m 1i 1
the behavieral seienees. Hillsdale, NJ: Lawrence Erlbaum Associates.

87

88

Denisi, A. S, Cafferty, T. P. & Meglino, B. M. (1984). Acognitive view of the
performance appraisal process: A model and research propositions.

WW2 33. 360-396
DohertY. M. L (1987). Wm
' i ' ' . Unpublished master’s thesis, Michigan State
University, East Lansing.

Epstein, S. (1980). The self-concept: A review and the proptgsal of an integrated

theory of personality. In E. Staub (Ed) Petsenality; aBsie issues an nu
eurrent resear r.eh Englewood Cliffs, N. J.: Prentice- Hall.

Erdelyi, M. H., & Appelbaum, G. A. (1973). Cognitive masking: The disruptive
effect of an emotional stimulus upon the perception of contiguous neutral

items. Won. 1. 560-61.

Feldman, J. M. (1981). Beyond attribution theory: Cognitive processes in
performance appraisal. l fA li h l 66, 127-148.

Feldman, J. M., & Hilterman, R. J. (1975). Stereotype attribution revisited: The
role of stimulus characteristics, racial attitude, and cognitive differentiation.

qurnal pf Persenality, 31, 1177-1188.

Fiske, S. T., & Taylor, S. E. (1984). SueiachguiuQu. Reading, MA: Addison-
Wesley.

Fong, G. T., & Markus, H. (1982). Self-schemas and judgments about others.
Socialmnitiou, 1, 191-204

Freud, S. (1956). Further remarks on the defense neuropsychoses. In Qalleeted

papers ef Sigmund Freud, Vol 1. London: Hogarth Press. (Original work
published 1924).

Geller, V., & Shaver, P. (1976). Cognitive consequences of self-awareness. Journal
pf Experimental Seeial Psyehglegy, 12, 99-108.

Gough, H. G., & Heilbrun, A. B. (1965). W. Palo
Alto, CA: Consulting Psychologists Press.

Greenwald, A. G. (1982). Ego task analysis: An integration of research on ego-
involvement and self-awareness. In A. Hastorf, & A. Isen (Eds.), Cagnitive

suciLlpsycmquy. New York: Elsevier.

Graesser, A. C., Woll, S. B., Kowalski, D. J ., and Smith, D. A. (1980). Memory for

typical and atypical actions in scripted activities. Jeurnal Qt Experimental
Psyehelegy; Human Learning and Memety, 6, 503-515.

89

Hall, G. S. (1898). Some aspects of the early sense of self. @eriean Jeurnal ef
Rwandan. 2. 351-395.

Hammer, W. C., Kim, J. S, Baird, I... & Bigoness, N. J. (1974). Race and sex as
determinants of ratings by potential employers 1n a simulated work sampling
task MW 52. 705-711

Heron, A. (1956) The effects of real- life motivation on questionnaire response.

humaLQLADnMBXQhQLQEX-flfi-LS

Higgins, E. T, & King, G. A. (1981) Accessibility of social constructs: Information
processing consequences of individual and contextual variability. In N.

Cantor & J Kihlstrom (EdS) WW
Hillsdale, N .J Lawrence Erlbaum Associates.

Higgins, E. T., King, G. A., & Mavin, G. H. (1982). Individual construct accessibility
and subjective impressions and recall. rn l r n li '

Em. i2. 35-47.

Higgins, E. T., Rholes, W. S., & Jones, C. R. (1977). Category accessibility and

impression formation. qurnal ef Experimental Seeial Psyehplpgy, 13, 141-
154.

Holmes, D. S. (1978). Projection as a defense mechanism. Psyehelegieal Bulletin,
86, 677-688.

Ilgen, D. R., Barnes-Farrell, J. L., & McKellin, D. B. (1986). Pedeguapeesatmg
aeeutaey. Paper presented and the let International Congress of Applied
Psychology, Jerusalem, Israel.

Ilgen, D. R., & F avero, J. L. (1985). Limits in generalization from psychological

research to performance appraisal processes. Aeauemy ef Management
Review, 10, 311-321.

Ilgen, D. R., & Feldman, J. M. (1983). Performance appraisal: A process

approach. In B. M. Staw (Ed.), WW Vol. 2.
Greenwich, CT: JAI Press.

J acoby, J. W., Chestnut, R. W., Weigl, K. C., & Fisher, W. A. (1976). Prepurchase
information acquisition: Description of a process methodology, research
paradigm, and pilot investigation. In B. B. Anderson (Ed.), Advanees jg

mugs-eaten. Vol.3.

James, L. R., & Brett, J. M. (1984). Mediators, moderators, and tests for mediation.

l f A 1i 1 62, 307-321.
James, W. (1915). Psycholegy, briefet epurse. New York: Holt.

90

Kelly, G. A. (1955). W. New York: Norton.

Kihlstrom, J., & Cantor, N. (1984). Mental representations of the self. Auvanees in
espeu'mental social psyehelogy, 11, 1-47.

Klinger, E. (1975). Consequences of commitment to and disengagement from

incentives. Psyehel egieal Review, &, 1-25.

Kuiper, N. A. (1981). Convergent evidence of the self as a prototype: The "inverted-

URT effect" for self and other judgments. n 1i ' 1P h 1
Bulletin, 1, 438-443.

Kuiper, N. A., & Derry, P. A. (1981). The self as a cognitive prototype: An
application to person perception and depression. In N. Cantor & J.
Kihlstrom (Eds.), n li ' i n n ' in r i . Hillsdale,
N. J .: Lawrence Erlbaum Associates.

Iandy, F. J ., & Farr, J. (1980). Performance rating. Esyehelegieal Bulletin, 81, 72-
107.

Iatham, G. P., & Wexley, K. N. (1977). Behavioral observation scales for
performance appraisal purposes. Persunnel Psyehelegy, 30, 255-268.

Lewicki, P. (1983). Self-image bias in person perception. Jeurnal of Persenality and
Seeial Psyehulugy, 46, 384-393.

Lewicki, P. (1984). Self-schema and social information processing. Jeurnal of
n li P h 1 _Z, 1177-1190.

Lord, R. G., & Foti, R. J. (1986). Schema theories, information processing, and
organizational behavior. In H. P. Sims & D. A. Gioia (Eds.), The thinking
mgauiaatiuu. San Francisco: J ossey-Bass.

Luria, A. R., & Vinogradova, O. S. (1959). Anobjective investigation of the
dynamics of semantic systems. MW 60, 89-105.

Markus, H. (1977). Self-schemata and processing information about the self.

WW3: 6378

Markus, H. (1983). Self-knowledge: An expanded view. Journal 6f Persenality, _5_1,
543-565.

Markus, H., Crane, M., Bernstein, S., & Siladi, M. (1982). Self-schemas and gender.
Jeurual 6f Persunality anu Seeial Psyehelegy, 62, 38-50.

Markus, H., & Nurius, P. (1986). Possible selves. W, 61,, 954-
969.

91

Markus, H., & Smith, J. (1981). The influence of self-schemata on the perception of

others. In N. Cantor & J. Kihlstrom (Eds) Wm
jutesagiup. Hillsdale, N ..:J Lawrence Erlbaum Associates.

Markus, H., Smith, J ., & Moreland, R. L. (1985). Role of the self-concept in the

perception of others. Jeurnal 6f Persenality anti Sega] Psyehelugy, 62, 1494-
1512.

Markus, H., & Wurf, E. (1987). The dynamic self-concept: A social psychological
perspective. Annual Review ef Psyehelegy, 38, 299-337.

Matte, W. E. (1982). An experimental investigatieu 6f infermatieu seareh in
pefioguameapptaigl. Unpublished doctoral dissertation, Purdue
University.

Mead, G. H. (1934). Mind If n i r m h 11 int f
behavierist. Chicago: University of Chicago Press.

Nisbett, R., & Ross, L. (1980). Human interferenee; Strategies and shurtcomings of
seeial judgment. Englewood Cliffs, N.J.: Prentice-Hall.

Nunnally, J. C. (1978). Psyehometrie theety. New York: McGraw-Hill.

Payne, J. W. (1976). Task complexity and contingent processing in decision making:

An information search and protocol analysis. Qrganizatienal Behavier and
Humau Perfurmanee, 16, 366-387.

Rogers, C. R. (1951). Client-sentereu therapy. Boston: Houghton-Mifflin.

Rogers, T. B. (1981). A model of the self as an aspect of the human information
processing system. In N. Cantor & J. Kihlstrom (Eds.), Persenality,

epguitiun, anti soeial interaetien. Hillsdale, N. J.: Lawrence Erlbaum

Associates.

Rogers, T. B., Kuiper, N. A., & Kirker, W. S. (1977). Self-reference and the

encoding of personal information. Juurn al at Persenality an nu Suei 1al
Psyehulogy, 36, 677- 688.

Rosch, E. (1978). Principles of categorization. In E. Rosch & B. B. Lloyd (Eds.),
Qegnitien and eategerization. Hillsdale, N. J .: Lawrence Erlbaum

Associates.

Schechtman, S. L. (1987). The effeets 6f appraisal putpuse auu pu’er kuewleuge:

Au infermatien preeessing perspeetive. Unpublished master’s thesis,
Michigan State University, East Lansing.

 

92

Schiffenbauer, A. (1974). Effect of observer’s emotional state on judgments of the

emotional state of others. ' P r ' i 11’ h l . 10.
31-35.

Schlenker, B. R. (1980). httppessitmauagemeut. Monterey, CA: Brooks/Cole.

Schlenker, B. R. (1985). Identity and self-identification. In B. R. Schlenker (Ed.),
Ihe_s_e_lf_aud_s_o_ei_al_li_fe. New York: McGraw-Hill.

Sechrest, L. B. (1968). Personal constructs and personal characteristics. Juurnal of
Individual Psyehulegy, 2_4,, 162-166.

Shiffrin, R. M., & Schneider, W. (1977). Controlled and automatic human
information processing: II. Perceptual learning, automatic attending, and a

general theOI'Y- EW- 8A- 127-190-

Shrauger, S. J ., & Patterson, M. B. (1974). Self-evaluation and the selection of
dimensions for evaluating others. Jeurnal 6f Petsonality, 62, 569-585.

Srull, T. K, & Wyer, R. S. (1979). The role of category accessibility on the
interpretation of information about persons: Some determinants and

implications. Jourual df Bersenality and Seeial Psyehology, 31, 1660-1672.

Srull, T. K., & Wyer, R. S. (1986). The role of chronic and temporary goals in social
information processing. In B. M. Sorrentino & E. T. Higgins (Eds.),

Handpddk 6f metivatien and eeguitien; Feundatieus 6f sueial behavidr.
New York: John Wiley & Sons.

Stewart, D., & Love, W. (1968). A general canonical correlation index.
Psyehelegieal Bulletiu, 19, 160-163.

Stryker, S. (1980). Sytubelie interaetidnism. Menlo Park, CA:

Benjamin/ Cummings.

Sullivan, H. S. (1953). The intetpersonal theety of psyehiatty. New York: Norton.

Taylor, S. B., & Crocker, J. (1981). Schematic bases of social information
processing. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds. ),S_Q_c__

W, Vol. 1. Hillsdale, N ..:J Lawrence21

Erlbaum Associates.

Triandis, H. C. (1976). Variatidns in blaek and white pereeptions 6f the social
enyitunment. Urbana, IL: University of Illinois Press.

Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging
frequency and probability. Cognitive Psyehelegy, 6, 207-232.

93

Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and
biases. Sg'euee, 86, 1124-1131.

Williams, K. J, DeNisi, A. S., Blencoe, A. G, & Cafferty, T. P. (19.85). The role of
appraisal purpose: Effects of purpose on information acquisition and

utilization WM 35
314-339.

Winer. BJ- (1971). WWW New York:
McGraw-Hill.

Wishner, J. (1960). Reanalysis of "impressions of personality." Psyeheldgieal
W. 91. 96-112.

Wyer, R. S., & Srull, T. K. (1980). The processing of social stimulus information: A
conceptual integration. In R. Hastie, T. M. Ostrom, E. B. Ebbesen, R. S.

Wyer, D. L. Hamilton, & D. E. Carlston (Eds), W
W Hillsdale N J.: Lawrence Erlbaum
Associates.

Wyer, R. S., & Srull, T. K. (1981). Category accessibility: Some theoretical and
empirical issues concerning the processing of social stimulus information. In
E. T. Higgins, C. Herman, and M. Zanna (Eds.), Seeial eognition; The
Qntarie symposium Vol. 1. Hillsdale, N. J .: Lawrence Erlbaum Associates.

 

APPENDICES

APPENDIX A

Consent Form

94

Performance Rating Study

This study is designed to assess how'people like yourself rate
other people who are performing a job that you know something
about. In this study, you will be asked to do a number of things.
First, you will be asked a few questions about yourself, including
some basic information about your experience as an employee for a
fast food restaurant. Second, you will be asked to describe those
areas of your job as a fast food employee in which you feel or felt
most competent. Third, you will be asked to rate your actual
performance on a number of tasks that you may have performed while
a fast food restaurant employee. Finally, you will be requested
to work on a computer simulation task in which you take on the role
of a restaurant manager who needs to make a performance rating
about one of your employees. In order to make this rating, you
will search for the information you need about this employee's
performance using the computer.

Your participation in this research is completely voluntary.
Therefore, you may discontinue your participation at any time;
there will be no penalty or recrimination. In addition, all
information you provide will be kept in the strictest confidence.
Your answers will be kept anonymous, and will be accessible only
to the experimenter, David McKellin. Finally, while your data may
be combined with that of others in a summary report, your data will
not be distributed in any way in which it can be identified
individually as yours.

Please acknowledge that you understand and agree to the terms of
this research by reading the statement below and then complete the
blanks at the bottom of the page.

We

I agree to participate in this Performance Rating Study. I
understand that I will be asked to fill out several questionnaires
regarding myself, including questionnaires about my background and
to evaluate my own performance as an employee in a fast food
restaurant. Furthermore, I understand that my participation is
voluntary, that I may discontinue my participation at any time
without recrimination, that my answers will be kept confidential,
and that my responses will not be distributed in any way in which
they can be identified as mine.

Signature Date

Print name

 

APPENDIX B

Background Questionnaire

95

Wail-on

How old are you?

 

years

What is your gender?

Female Male

 

For which fast-food restaurant do/did you work?

Burger King _____ McDonalds Wendys

 

 

How many months have you/did you work for this company?

._____ Honths

Are you currently working for this company?

Yes No

 

 

now many months has it been since you worked for this company?
(If you are currently working as an employee in a fast food
restaurant, please enter 0.)

Months

 

96

W.

The remainder’of this-questionnaire asks questions about particular
duties or tasks in Fast Food Restaurants. We call these stations.
Please answer each question about the various stations in these
restaurants.

7.

U) bu N PO

Using the following scale, please indicate how much experience
you have had performing the duties associated with each of
stations listed.

- No experience at all.

- Have performed some of the duties, but still one of the least
experienced employees in the store.

- More experience than some, but less experienced than most other
employees. '

- A moderate or average amount of experience

- More experienced than most, but not as experienced as some
other employees.

- One of the most experienced employees in the store.

Dining room maintenance
Counter service

Outdoor maintenance
Sandwich preparation

Restroom maintenance

 

French fry production

What percentage of your time per week, on average, did you spend
working at each station listed below when you terminated your
employment at the restaurant? (If you are presently working at
the restaurant, what percentage, on average, do you currently
spend working at each station?) On each line record a numeral
between zero and 100 to indicate the percent of time spent
working at each station. The sum of all stations should equal
100%.

Dining room maintenance

Counter service

Outdoor maintenance
Sandwich preparation

Restroom maintenance

 

French fry production

100% - Total

97

9.ldsted below are six job stations.

new

, think of the TEN TASKS YOU FEEL YOU

PERFORMED.BEST while working in the rostaurant (e.g. filling
customer's orders accurately, dressing sandwiches correctly).
Then, please list each of these ten tasks under their
corresponding station. That is, list where you did these tasks.
When you complete this measure some dimensions may have nothing
listed and some may have more than one task listed. It is
important, however, to list 10 AND ONLY 10 TOTAL on this page.

W

7:” -—- 4.7.17:-

 

APPENDIX C

Computer Task Tutorial

7“: «...... A a“

98

W E L C O H E

This exercise is a simulation of a particular type of decision faced by
supervisory personnel in most organizations. Today, you will be taking the
role of a manager in a fast food restaurant and will be required to make a
series of rating decisions regarding the evaluation of one of your
subordinate hourly employees.

Organizations utilize performance evaluation information for a variety

of different purposes. One use of performance ratings is to make decisions
regarding pay increases and promotion. Another is to provide feedback to an
employee about how well he or she is performing his or her job.

Today, you will be required to evaluate the performance of one of your
part-time hourly employees. It is important to be very careful in making
accurate ratings about your employee's performance. Bad performance rating
decisions reflect poorly on the manager who is making the rating.

PLEASE PRESS THE RETURN BUTTON FOR FURTHER EXPLANATION

99

When a supervisor looks to rate the performance of his or her employees,

he or she usually doesn't evaluate the employee‘s overall performance in
general, but rather makes judgments about the quality of the worker's
accomplishments along each of a number of different job dimensions.

As a restaurant manager, your task will be to find out about your employee's
performance along each of six job dimensions, and then to rate the quality
of that performance within each dimension.

If you have any questions, please ask the experimenter for assistance. If
you do not have any questions, press the RETURN button and you will receive
more specific instructions about your task.

100

To aid you in searching for information about your employee's performance,
you will be presented with two lists. The first list contains six job
dimensions along which you will be rating your employee, and is labelled
DIMENSIONS. The second list, which contains descriptions of your employee's
job behaviors on various tasks within a job dimension is labelled
ALTERNATIVES. For' example, if’ you. were rating the performance of a
secretary, you might see a screen such as:

DIMENSIONS ALTERNATIVES
1=TYPING l-BEHAVIOR A
2=PHONE HANNERS 2-BEHAVIOR B
3-FILING 3=BEHAVIOR C
4-SCHEDULING 4=BEHAVIOR D

As you can see, each alternative and each dimension are identified by a
number. To begin searching for information, you will be asked two questions:
(1) the dimension number about which you would like information and (2) the
alternative number about which you would like to receive information. Using
the number keys on the row above the typewriter keypad, simply type the
number corresponding to the dimension you would like and then type the number
corresponding to the alternative you would like.

PRESS THE RETURN BUTTON TO CONTINUE

101

CONFUSED? Let's go through the evaluation process in detail.

DIMENSIONS ALTERNATIVES
I-TYPING l'BEHAVIOR A
Z'PHONE MANNERS ZIBEHAVIOR B
3=FILING 3-BEHAVIOR C
48$CHEDULING 4-BEHAVIOR D

To begin the search process, you will choose one dimension and one
alternative of information describing that dimension. You will continue this
procedure until you have enough information to evaluate each employee's
performance, or until you have viewed two-thirds of the possible information.
At that time, you will be provided with a rating scale, ranging from 1 to 9
(1 a exceptionally poor performer; 9 - exceptionally good performer). Using
this scale, you will be asked to provide a rating for each individual on each
dimension. .

PRESS THE RETURN BUTTON TO CONTINUE

—--—-“

102

To see how this procedure works, let's begin with the following lists:

DIMENSIONS ALTERNATIVES
I-TYPING 1-BEHAVIOR A
2=PHONE NANNERS ZIBEHAVIOR B
3=FILING 3=BEHAVIOR C
4-SCHEDULING 4=BEHAVIOR D

The following message will appear below the dimensions and alternatives:

ENTER THE NO. OF THE DIMENSION AND HIT RETURN ?

ENTER THE NO. OF THE ALTERNATIVE AND HIT RETURN ?

Let's assume that you are interested in TYPING BEHAVIOR C. You would press
-1- for TYPING, the RETURN button, and then -3- for BEHAVIOR C and the RETURN
button. The present screen*will disappear and the requested information will
be shown on the next screen as follows:

PRESS THE RETURN BUTTON TO CONTINUE

103

Employee misspells every fourth word in a report which goes to the company
president.

PRESS THE RETURN BUTTON TO CONTINUE

_..-_A— ..‘l‘

104

At this point, the computer will display the following message:

ENTER 1: IF YOU NEED MORE INFORMATION

2: IF YOU ARE READY TO MAKE THE FINAL RATINGS

Let's assume that you are not ready to make a the final ratings and would
like more information. You would press -1- and the RETURN button. The
computer will then print the original menu on the next screen.

PRESS THE RETURN BUTTON TO CONTINUE

 

105

 

DIMENSIONS ALTERNATIVES
18TYPING l-BEHAVIOR A
ZIPHONE MANNERS 2-BEHAVIOR B
3=FILING 3-BEHAVIOR C
4-SCHEDULING 4tBEHAVIOR D
ENTER THE NO. OF THE ALTERNATIVE AND HIT RETURN ?
ENTER THE NO. OF THE DIMENSION AND HIT RETURN ?

Now let's suppose you want to know about the PHONE BANNERS used in BEHAVIOR
D. You would type in a -2- for PHONE NANNERS and a -4- for BEHAVIOR D.

PRESS THE RETURN BUTTON TO CONTINUE

106

Now the computer displays the following message:

The employee answered the phone with the department's name and gave his name.

panss THE RETURN BUTTON TO CONTINUE

107

At this point, the computer will display the following message:

ENTER 1: IF YOU NEED MORE INFORMATION

2: IF YOU ARE READY TO MAKE THE FINAL RATINGS

Again, let's assume that you are not ready to make the final promotion
ratings. After pressing -1- for more information, the computer will return
to the original menu on the next screen.

PRESS THE RETURN BUTTON TO CONTINUE

108

DIMENSIONS ALTERNATIVES
18TYPING I-BEHAVIOR A
Z-PHONE MANNERS ZBBEHAVIOR B
3=FILING 3iBEHAVIOR C
4=SCHEDULING 4=BEHAVIOR D
ENTER THE NO. OF THE DIMENSION AND HIT RETURN ?
ENTER THE NO. OF THE ALTERNATIVE AND HIT RETURN ?

Now let's assume that you want to see SCHEDULING performance on BEHAVIOR C.

You would type a -4- for SCHEDULING and a -3- for BEHAVIOR C.

PRESS THE RETURN BUTTON TO CONTINUE

109

Employee schedules two meetings in the same conference room at the same time.

PRESS THE RETURN BUTTON TO CONTINUE

110

VT”!!! '
A '

-TF5

Again, the computer will display the following message at this time:

ENTER 1: IF YOU NEED MORE INFORMATION

2: IF YOU ARE READY TO MAKE THE FINAL RATINGS

At this point, let's assume that you have searched a sufficient number of
different dimensions of performance for each employee and are ready to make
a rating. You would type a -2- and the RETURN button.

PRESS THE RETURN BUTTON TO CONTINUE

111

The computer will now display the following message:

ENTER <N> IF YOU ARE NOT READY TO COMPLETE THE RATINGS
ENTER <Y> IF YOU ARE READY To COMPLETE THE RATINGS
Since you are ready to complete the ratings, you would press the -Y- key and

the RETURN button. The present screen will disappear and a 9-point
scale for making the ratings will appear on the next screen.

PRESS THE RETURN BUTTON TO CONTINUE

 

112

The computer will now display the following message:

On the following 9-point scale, please rate the employee's performance on
each job dimension by entering the appropriate number from 1 to 9 after the
colon. NOTE: Once ratings are entered, they cannot be changed.

 

I ------ I ------ I ------ I I I ------ I ------ I ------ I
1 2 3 4 5 6 7 8 9
EXCEPTIONALLY NEITHER GOOD EXCEPTIONALLY
POOR PERFORMER NOR POOR PERFORMER GOOD PERFORMER
TYPING :
PHONE MANNERS :
FILING :
SCHEDULING :

Following the cursor's prompt (e.g., the colon), you would rate the
employee's performance by entering the appropriate number along the 9-point
scale after the colon.

PRESS THE RETURN BUTTON TO CONTINUE

113

At this point, the computer will summarize the results of your ratings as
follows:

DIMENSION RATING
TYPING :
PHONE MANNERS :7
FILING :5
SCHEDULING :

PRESS THE RETURN BUTTON TO CONTINUE

 

114

 

Are you ready to continue and make your own responses? If you are ready,
press the RETURN button to continue. If you are not ready, ask the
experimenter to clarify an questions you may have. GOOD LUCK!

PRESS THE RETURN BUTTON TO CONTINUE

 

APPENDIX D

Computer Search Practice Task

115

Now that you are familiar with the search procedure, you will be given an
opportunity to practice your new skills prior to conducting the actual
performance evaluation. For this practice task, you will evaluating the
performance of a university professor. This professor's performance will be
described along 4 performance dimensions. You can search for as little or
as much information as necessary in making your evaluation. The dimensions
and alternatives are as follows:

DIMENSIONS ALTERNATIVES
1=LECTURING 1=BEHAVIOR A
2=ANSWERING QUESTIONS 2-BEHAVIOR B
BBADVISING 3=BEHAVIOR C
4=EXAM WRITING 4'BEHAVIOR D

Remember to choose one dimension and one alternative at a time. Type in the
number corresponding the desired dimension, hit RETURN, and then type the
number corresponding to the desired alternative and hit RETURN. Continue
this procedure until you are ready to make the performance ratings for the
professor's performance.

PRESS THE RETURN BUTTON TO CONTINUE

116

 

DIMENSION ALTERNATIVE
1: LECTURING 1: BEHAVIOR A
2: ANSWERING QUESTIONS 2: BEHAVIOR B
3: OFFICE HOURS 3: BEHAVIOR C
4: EXAMINATIONS 4: BEHAVIOR D

ENTER NO. OF DIMENSION FROM 1 TO 4 THEN RETURN ?
ENTER NO. OF ALTERNATIVE FROM 1 TO 4 THEN RETURN

 

 

117

Professor explains a complex concept so that class understands it

ENTER 1: IF YOU NEED MORE INFORMATION
2:

IF YOU ARE READY TO MAKE THE FINAL RATINGS
., .

 

118

On the following 9-point scale, please rate the employee's performance on
each job dimension by entering the appropriate number from 1 to 9 after the
colon. NOTE: Once ratings are entered, they cannot be changed.

I ------ I ------ I ------ I ------ I ------ I ------ I ------ I ------ I
1 2 3 4 5 6 7 a 9
EXCEPTIONALLY NEITHER coon EXCEPTIONALLY
POOR PERFORMER non poon PERFORMER coon PERFORMER
TYPING : '
PHONE MANNERS :
FILING :

SCHEDULING

APPENDIX E

Computer Task Introduction Screen

119

Now that you have done an example, you should be ready to begin the
experiment. You will be presented with a list of six job dimensions
(DIMENSIONS) related to the work of hourly employee in a fast food restaurant
along which you are to rate your subordinate. You also will be presented
with list of seven job behaviors (ALTERNATIVES). Your task is to examine as
little or as much information as necessary to provide an accurate rating of
your employee's performance on each job dimension. You will be able to
search for 21 PIECES of information about your employees performance. This
should allow you to collect plenty of information about your employee's
performance, but IT IS IMPORTANT THAT YOU CHOOSE INFORMATION CAREFULLY, so
that you can make the best ratings possible. The dimensions along which you
will be evaluating your employee are:

1. French Fry Production
2. Restroom Maintenance

3. Dining Room Maintenance
4. Counter Service

5. Sandwich Preparation

6. Outside Maintenance

If you have any questions, please ask the experimenter at this time.
Otherwise, press the RETURN BUTTON to begin the experiment.

 

APPENDIX F

Computer Task Behavioral Descriptions

III

 

120

WWW
Alf. Marion

5:.“ __ ..:-:53

1 Employee shakes fry basket after 1 minute of cooking (-)

2 Employee holds only the fry bag currently being filled, even
during rush periods (-)

3 Employee discards cooked fries held in the bagging station longer
than 7 minutes (+)

4 Employee fills bags or cartons of fries so that they are just full

5 Employee salts fries using front-to-back motion (+)

6 Employee rotates fries first in - first out in bagging station (+)

7 Employee)does not drain shortening from basket before dumping

ries -

:[ll;“l -Dll 1111::
Alf-Maxim:

1 Employee cleans ash trays and replaces as necessary (+)

2 Employee misses trash on some tables (-)

3 Employee assists customers by helping them to carry a large order
to their table and by getting them a high chair (+)

4 Employee leaves a small spill alone, intending to mop it up later

5 Employee forgets to check for smudges on doors and windows (-)

6 Employee cleans table tops and sides, high chairs, and door

handles using cloth and sanitizer solution (+)

7 Employee spot sweeps only area directly in front of counter (-)

121

MW

51:3 lQh.B£h§!19I

Employee notices paper products are low, but determines that they
can wait until the next shift comes on (-)

Employee checks and refills soap supplies (+)

Employee leaves streaks on mirror and wash basin after cleaning

(')

Employee checks washroom only when there is nothing else to do

(')

Employee keeps walls and partitions clean and free from smudges
and graffiti (+)

Employee forgets to check toilets to see if they are clean (-)

Employee empties waste baskets and makes sure they are clean (+)

W
Maxie;
Employee misses dirt on outside of windows (-)
Employee pushes down trash in compactor (+)

Employee forgets to empty and clean several outside trash bins

(')

Employee cleans drive-thru sign and equipment and checks to see if
it is functioning properly (+)

Employee closes the gate to the trash corral (+)

Employee picks up all trash on the sidewalks, grass, and street
gutters (+)

Employee does not notice gum left on sidewalks (-)

 

122

WWW

Alt-W

Employee assembles sandwiches correctly (+)

Employee forgets to put two condiments on sandwich (-)

Employee makes sure that cooking area is clean and sanitized (+)
Employee cooks meat correctly (+)

Employee insures that sandwich buns are toasted correctly (+)
Employee runs out of raw products (meat, dressings, etc.) (-)

Employee forgets to cleans hands before taking over'cooking duties

(')

W

Employee takes customer's order, suggesting any items which may be
missing (+) .

Employee gives customer too much change (-)

Employee gives customer correct number of condiment packets for
order (+)

Employee is slow in assembling order (-)
Employee greets customer, saying ”Hi, how are you doing?" (-)

Employee forgets to check the quality of products before giving
them to the customer (-)

Employee delivers the correct products to the customer (+)

.4‘
F'

 

APPENDIX G

Self-rated Performance Questionnaire

 

123

WWW

Below are a list of activities that are often described as part of the
job duties of persons working in various positions in fast food
restaurants. Using the scale below, rate your honest opinion about the
frequency with which you met the restaurant's standards for good
performance of the activity that is described. If you never worked the
station responsible for the particular activity listed, please enter 0
on the blank.

 

RATING SCALE

Regularly performed below the standard.

Occasionally performed below standard and performed at the

standard the rest of the time.

Performed at the standard.

:- Occasionally exceeded the standard but usually performed at the
standard.

- Performed above the standard about half the time and at the
standard the rest of the time.

- Frequently performed above the standard and at the standard the
rest of the time.

. Almost always exceeded the standard.

- So consistently exceeded the standard that I stood out to both

supervisors and peers as an exceptionally good performer of

this task.

bid Mia
I

010)

9‘1

1. Suggesting additional items when taking customers'
orders

2. Shaking fry baskets after the fries had been cooking
for 30 seconds

3. Cleaning and replacing ash trays when necessary _____
4. Keeping restroom paper products fully stocked
5. Checking and cleaning the outside of windows
6. Assembling sandwiches correctly

7. Making appropriate change for customers

8. Holding the correct number of fry bags when
bagging french fries

9. Clearing trash from unoccupied tables in the
dining room

10. Checking restroom soap supplies, and refilling ‘ _____
them when necessary

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23C

24.

25.

124

RATING SCALE

Regularly performed below the standard.

Nia
I

standard the rest of the time.
- Performed at the standard.

fit»
I

standard.

standard the rest of the time.

at U)
I

rest of the time.
- Almost always exceeded the standard.

fix)
I

this task.
Making sure that trash in the trash compactor was
crushed
Applying correct condiments to sandwiches

Giving customers the appropriate number of
condiment packets for their orders

Discarding cooked fries held in the bagging
station longer than 7 minutes

Helping customers find high chairs when they
need them

Cleaning restroom mirrors and wash basins
Emptying and cleaning all outside trash bins
Keeping the cooking area clean and sanitized
Assembling customers' orders quickly

Filling fry bags or cartons so they are just
full

Mopping up spills in the dining room promptly
Checking restrooms at the required time intervals

Cleaning and checking the operation of the
drive-thru equipment

Cooking meat correctly

Welcoming customers using the appropriate
greeting

Occasionally performed below standard and performed at the

Occasionally exceeded the standard.but usually performed at the
Performed above the standard about half the time and at the
- Frequently performed above the standard and at the standard the

So consistently exceeded the standard that I stood out to both
supervisors and peers as an exceptionally good performer of

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

125

RATING SCALE

Regularly performed below the standard.
Occasionally performed below standard and performed at the
standard the rest of the time.
Performed at the standard.
Occasionally exceeded the standard.but usually performed at the
standard.
- Performed above the standard about half the time and at the
standard the rest of the time.
- Frequently performed above the standard and at the standard the
rest of the time.
- Almost always exceeded the standard.
- So consistently exceeded the standard that I stood out to both
supervisors and peers as an exceptionally good performer of
this task.

ht» Mid
I

ms) 05 U)

Salting fries using a front-to-back motion

Removing all spots and smudges from doors and
windows

Keeping restroom walls and partitions clean and
free from smudges and graffiti

Keeping the gates to the trash corral/area
closed

Toasting sandwich buns correctly

Checking the quality of products before giving
them to customers

Rotating packaged fries first in - first out
in the bagging station

Cleaning table tops and sides, high chairs, and
door handles using cloth and soapy water or sanitizer
solution

Checking to make sure toilets are clean

Picking up all trash on sidewalks, grass, and
street gutters

Making sure that sufficient supplies of raw
products are maintained

Delivering the correct products to customers

Draining shortening from fry baskets before ‘ _____
dumping fries into the bagging area

 

39.

40.

41.

42.

126

RATING SCALE

Regularly performed below the standard.
Occasionally performed below standard and performed at the
standard the rest of the time.
Performed at the standard.
Occasionally exceeded the standard but usually performed at the
standard.
- Performed above the standard about half the time and at the
standard the rest of the time.
- Frequently performed above the standard and at the standard the
rest of the time.
- Almost always exceeded the standard.
- So consistently exceeded the standard that I stood out to both
supervisors and peers as an exceptionally good performer of
this task.

Gd 0) U) this) NH

Spot sweeping the dining room floor

Emptying restroom waste baskets and making sure they
are clean

Removing gum from the sidewalks around the building

Washing my hands before taking over cooking
responsibilities

 

APPENDIX H

Dimension Importance Rating Questionnaire

 

127

‘ W

In this part of the study, please use the scale below to indicate how
ggigiggl each task is for the successful job performance of fast food
employees. Please rate the importance of these tasks gygn if you never
w- .’-_ f , - 'l ‘ "l: .- 9 - - ‘-

n2XQI__QIB2Q_IQI.§.£A§§.IQQQ.I§§£§QI§H§- NOTICE THAT THESE RATINGS USE
A 7-POINT, INSTEAD OF AN B-POINT, RATING SCALE. Rate each item by

placing the numeral int the blank that best describes your beliefs.
RATING SCALE

1 - Task is not critical at all to performing successfully as a fast
food employee.

2 a Task is remotely critical to performing successfully as a fast food
employee, but can be considered one of the best employees even if
performance on this task is substandard.

3 - Task is somewhat critical to performing successfully as a fast food
employee, but can be considered a good employee even if performance
is substandard.

4 a Task is important in performing successfully as a fast food
employee: an employee who did not perform this task successfully
would only be considered an average employee.

5 a Task is very important in performing successfully as a fast food
employee: an employee who did not perform this task successfully
would keep others from getting their work done properly.

6 - Task is crucial in performing successfully as a fast food employee;
an employee who failed to perform this task successfully would be
reprimanded, or possibly put on probation.

7 = This task is so critical that an employee who failed to perform it
successfully would probably be dismissed.

l. Suggesting additional items when taking customers'
orders

2. Shaking fry baskets after the fries had been cooking
for 30 seconds

3. Cleaning and replacing ash trays when necessary
4. Keeping restroom paper products fully stocked
5. Checking and cleaning the outside of windows

6. Assembling sandwiches correctly

7. Making appropriate change for customers

8. Holding the correct number of fry bags when
bagging french fries

9. Clearing trash from unoccupied tables in the
dining room

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

128

RATING SCALE

Task is not critical at all to performing successfully as a fast
food employee.

Task is remotely critical to performing successfully as a fast food
employee, but can be considered one of the best employees even if
performance on this task is substandard.

Task is somewhat critical to performing successfully as a fast food
employee, but can be considered a good employee even if performance
is substandard. ,

Task is important in performing successfully as a fast food
employee; an employee who did not perform this task successfully
would only be considered an average employee.

Task is very important in performing successfully as a fast food
employee: an employee who did not perform this task successfully
would keep others from getting their work done properly.

Task is crucial in performing successfully as a fast food employee:
an employee who failed to perform this task successfully would be
reprimanded, or possibly put on probation.

This task is so critical that an employee who failed to perform it
successfully would probably be dismissed.

Checking restroom soap supplies, and refilling
them when necessary

Making sure that trash in the trash compactor was
crushed

Applying correct condiments to sandwiches

Giving customers the appropriate number of
condiment packets for their orders

Discarding cooked fries held in the bagging
station longer than 7 minutes

Helping customers find high chairs when they
need them

Cleaning restroom mirrors and wash basins
Emptying and cleaning all outside trash bins
Keeping the cooking area clean and sanitized
Assembling customers' orders quickly

Filling fry bags or cartons so they are just
full

Hopping up spills in the dining room promptly

Checking restrooms at the required time intervals

 

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

129

RATING SCALE

Task is not critical at all to performing successfully as a fast
food employee.

Task is remotely critical to performing successfully as a fast food
employee, but can be considered one of the best employees even if
performance on this task is substandard.

Task is somewhat critical to performing successfully as a fast food
employee, but can be considered.aigood employee even if performance
is substandard.

Task is important in performing successfully as a fast food
employee: an employee who did not perform this task successfully
would only be considered an average employee.

Task is very important in performing successfully as a fast food
employee: an employee who did not perform this task successfully
would keep others from getting their work done properly.

Task is crucial in performing successfully as a fast food employee;
an employee who failed to perform this task successfully would be
reprimanded, or possibly put on probation.

This task is so critical that an employee who failed to perform it
successfully would probably be dismissed.

Cleaning and checking the operation of the
drive-thru equipment

Cooking meat correctly

Welcoming customers using the appropriate
greeting

Salting fries using a front-to-back motion

Removing all spots and smudges from doors and
windows

Keeping restroom walls and partitions clean and
free from smudges and graffiti

Keeping the gates to the trash corral/area
closed

Toasting sandwich buns correctly

Checking the quality of products before giving
them to customers

Rotating packaged fries first in - first out
in the bagging station

Cleaning table tops and sides, high chairs, and
door handles using cloth and soapy water or
sanitizer solution

 

130

RATING SCALE

Task is not critical at all to performing successfully as a fast
food employee.

Task is remotely critical to performing successfully as a fast food
employee, but can be considered one of the best employees even if
performance on this task is substandard.

Task is somewhat critical to performing successfully as a fast food
employee, but can be considered.a good employee even if performance
is substandard.

4 - Task is important in performing successfully as a fast food
employee; an employee who did not perform this task successfully
would only be considered an average employee.

5 - Task is very important in performing successfully as a fast food
employee: an employee who did not perform this task successfully
would keep others from getting their work done properly.

Task is crucial in performing successfully as a fast food employee:
an employee who failed to perform this task successfully would be
reprimanded, or possibly put on probation.

7 = This task is so critical that an employee who failed to perform it

successfully would probably be dismissed.

H
II

N
I

u
I

0
I

34. Checking to make sure toilets are clean _____

35. Picking up all trash on sidewalks, grass, and
street gutters

36. Making sure that sufficient supplies of raw
products are maintained

37. Delivering the correct products to customers

38. Draining shortening from fry baskets before
dumping fries into the bagging area

39. Spot sweeping the dining room floor

40. Emptying restroom waste baskets and making sure they
are clean

41. Removing gum from the sidewalks around the building

42. Washing my hands before taking over cooking
responsibilities

 

APPENDIX I

Zero-Order Correlations Among Measures

 

131

“2:380 03-:

888.5... .238 .8
3.33. 3.53 .m—
38. 8.5.. .8
sites pea-c.3eu

0.88353. soot:- .~—
Steel—see 5.3 .3
eussCeuc-sx 93:8 .0—
83...» .353 .9
Go. .550 .0

'0‘; s.u.un 25.05:

3.33 3:58 .
s8- 838 .~
8.3.893 sue..-

 

Jam-ll

-. .... 3. 2.. 8. e. 8. 3.. 3.. s. 3. 8. 8. 8. 8. a.
S. 8.. 8.. 8.. 2.... a. 2.. 3. 2. 8.. 8.. 8. 3. 9.
.. 8.. 8. a... 8. 8. n. 8. a. s. -. a. 3. 8.
. 8 8.. 8 8.. c. 3. 8.. 8. 8.. a: 8. 8. 8:38.. 558: .9
-. 8. 8. 8.. 8. 8. on. s. a. .... e. :.
.. 8. E: 8.. 8.. 8. 8. 3. 8.. 8. 2.
. a: 8. a... 8. a. 8. 3.. 8. 8.
. 8.. 8. 2. 3.. 8.. z. 8. 2.
a: 8. 8.. 8.. n... 8. s.
.. a. 2. :. ...». a. 3. 8.888.. 5.3.... .a
.. 8. 8. 8. a. 8. 88.55... 82.8.. ....
.. 8. 8.. 3. 8. 8.8.8312; .s
-. 8.. 8. n. 888.5... .238 J
.. 8. S. n
.. S.
.. 28.8.9.5 .9 ...-c... ..
finisasflizq—Iiooaonsnu.

 

DGqulflO: 0:0: mCOwUflHQNHOU HGGHOIOHON

132

 

8.228. .3...

8.828... s... .88... ..n
85:35... sofa-es .on
5.3.89... 5.2.3.3 .8

.8232... .288 .8
cotton .3258 .sn
88 2...£° .8
mesons: 38... .3

82863. t. gas-... .3.
8.8.88.8. 88.8. ......
8:21.... ......Psu .8

.8235... .288 .8
cutie» .353 ..n
..8. 8...... .8

ens-tone. 8.9.33

ens-.83.! 803:. .3

8.8.8.... £0.35... .2

 

8.. .. 8.. 2. 8. 8. 8.. 2.. 2. 8. 8. 2.. 8. 8.. 8. 8.
8.. 2.. 8. 8.. s. 8. 8.. 8.. 8. 2. 8.. 8.. 8. 8.. 8. 8.
8. 8.. 8. 8. 8. 8. 8. 8. 2.. 8. 2. 8. 8. 8. 2. 8.
8. 8. 8.. 8. 8. 8. 8. 8. 2.. 2. 8.. 8. 2. .. 8. 8.
8.. 8. 8. 2.. 2. 2.. 8.. 8. 2.. 8. 2.. 8. 2. 8. 8. 8.
8.. 8.. 8. 8. 8. s. 2.. 2.. 8. 8. 2.. 8.. 8. 8.. 8. 2.
8. 8. 8. 8. 8. 8.. 8. 2.. 8. 8. 2. 2. 8. 8. 8. 2.
8. 8. 3. 8. 2. 8.. 8. 8.. 2. 8. 8. 2. 8. 8. 2. 8.
2. 2. 8. 8. 8. 8. 8. 2.- 8. 8. 2. 8. 2. 2.. 8. 8.
8. 8. 8. 8. 8. 2.. 8. 8.. 8. 8. 2. 8. 2. 8. 2. 8.
8. 8. 8. 8. 2. 8.. 8. 2.. 2. 2. 8. 8. 8. 8.. 8. 8.
R. 3. en. 8. pm. 8. -. 2...- 3. S. 2. S. 2. ~93 Z. 3.
8. 8. 8. 8. 2. 8.. 8.- 8. 2.. 8. 8. 8. 8. 8. 8. 8. 8:38.. 5...... .2
2. 8. 8. 8.. 8. 8.. 8. 2.. :. 8. z. 8. 8. 8. 8. 2.
8. :. 8. 2. 8. 8. 8. 2.. 2.. s. 8. 8. 8. 8.. 8. 8.
3.233 unsafe?! pea-......“
AW: 8 8 2 2 2 o o k a n s n 8 .

eaflofll

 

133

ggi—.350 U—Dluv

 

 

S. .... .8.

IL

3.vn.o~.u..

 

 

 

3.. 3.. 9.... 3. so. 2. S. o... 3.- 3.. 3.. 3.. 3.. ~n.. 3.. 3.. ...-8.3V ataxia 8:; as: .3
a. 3.. 3.. 3. 3. 3. 3.. mm: 3.. 3.. no. on. on. 3.. 2.. 3.. no» .9
3.. ~... 3.. 3. 2. 3. 2. 9.... 3. 3.. 3.. 3. 3.. 3..... 8.. 3.. a: .3
8.8...» 88:88..
3.. 3. 8. 2. u... 3.. 3.. 8.. S. 2. 2. 3.. 2.. 3. 3.. 3.. 5.8.28... r. :98... .9
3. no. 8. 2.. .. n... .. 2. 3.. n... 3.. 3. 3. 3. 3.. 2. 02.8.35! [838. ......
3. 5.. 2.. on. 3. 3. fl. .. .. w... 3.. 3. 3. 3.. 2.. 3.. 8..-.30... 331.com .5
3. 3.. 3.. 2. 3. n3. 3. 3.. 3.. 3.. 3. 3. 3. 3.. 3.. 3. cue-58...: {.33 .3
m... an. 2. s»... 3.. n: 3.. 3. 3.. 3. 3.. 3.. n... no. 2. 2. 3.»..8 .859... in
S. 3.. S. 3.. nu. 3. 3.. 2.. on. 3.. n3. 3.. 3. 3.. 3.. 3.. 8:235... [8.. 3...... .on
83.3.... .o 3.... 88 .5. 8.8.8.? .8...»
3.. 8. 3.. 3.. 3.. 3.. 8. 3. 3.. 3. 3.. .. 3.. 8. 3.. 2.. 8.538.: s... ..:... .sn
3. 3.. no. 3. 3. 3. 3. 3.. 3.. 3. 3. 3. 3.. 3.. 3.. S. 08.2.85... 88.33. .3
3. 3.. .. 3.. 3. 3.. 8.. 3. 3. 8.. 8.. 8. 3.. 3.. 3.. ...... 8..-.30.... ..uifllu .nn
2.. n... 3. 3.. 8. 3.. 8. 3. 3. 3.. 3. 3. 3.. 3. n3. 3. 8..-58...: 8..-86 .3
3. 3.. 3.. no. 3.. 3.. 9.... o... 3. 3.. 3. 3.. 3.. 3. no. no. 8.»..8 .358 .3
3. 3.. 8. 3.. 3.. 3. ~.. 3.. 3.. 3. 3.. 3. 2. 3.. w... 3. l8. 3...... .~n
3.3 ..:-5.83 not... 9...... 8.8.3...

8 8 8 8 8 8 3 o 8 ~ 8 ... . n 8 . .88....

 

134

 

53:80 03.:

an. mo. an. km: 3.. 3.. 3.. 3.. 8.. 50.. 3.. 3.. «a. 3.. 53%: P: £9.35 .:
-.. 8.. 2.. a. .83 8. 8.. 3. 3.. S. 3.. 3.. 8.. mustangs: 895:: .3

.N. —N. 3.. n... 5.- me. —~.. 50.. 2.. 8. o... No. 8:05.59... total!” .24.

8.. on... 3. 2.. no. 3.. so. 3.. 3. 8.. 3.. toe-€350: .233 .8

3 8 8 8 a... 8 up 3.. 5 8 002C.» 53.590 .5...

3.. hp. 3.. 3. 8 8 an: n— 3 '08 .559 .8

32.9.00. 39. .a .0:

3 3. E n. S. 3. on. .n. 8:269: P: coco: .3
«a. 8 or: 8. «N. on. 3. coin-«£3. snob... .c~

mm. as. 3 en. 8. 2. 539.80.... :2: .nm

.. mo. 5. on. on. 0'. 09363508 {.33 .NN

.. no. on. ow. a. cotton 3.53 ..N

.. 5. on. —~. .8- !EE .8

8:383. 5.2353

on. «n. 8:262; >55 £93.: .op

94.. DES-ax 80.5-0- .2.

8:23.... .63....» .t

out-Fatwa tau-...:om

 

 

a flijlaljldll; : a «73 r all

 

135

all: 30 0.3:

 

 

 

3. 3. 3.. 3.. 3.. 3. 3. 3. 3. 3. 3. 5. 3.. 3.. 3.. $033. 3.5-u 8:3 on: .3
3. 3. 3. 3. 3.. 3.. 3.. 3. 3. 3.. 3.. 3.. 3.. 3.. 3. no» .9
3. 3.. 3.. 3. 3.. 3.. 3. 3. 3. 3. 3 3. 3.. 3. 3.. 00¢ .3
8.3.3) 2.183

3. 3 . 3. 3. 3.. E. 3.. 3. 3. 3. 3. 3. 3.. 3.. 3. 53819.; P: ..:... .3
3. 3.. 3. 3.. 3. 3.. 3. so. 3. 3. 3 . 3. 3 . 3. 3. 35:358. [8.58- .3
3. 3. 3. 3. ow. 3.. 3. 3. kw. 3. am. 3. 3.. 3.- 3.- cot-LIZ; :03?!» .3
3. 3.. 3 3.. 3. 3. 3. 3. 3. 3. 3. 3 3 . 3. 3. 85:358. .383 .3
3.. 3. 3.. 3. 5. 3. 3.. 3. 3.. 3.. 3.. 3.- 3 3.. 3.. 03>...» 5853 .on
3. 3 8.. n... 3.. rd. 3.. 3. 3. 3. 3. 3. 3.. 3. 3.. 8..-:35... I8: 3.:3 .3
833.52: 3 30... as .3. 5.9!... 30.8

3.. 3 . 3. 3.. ow. 3. 3. 3.. 3. 3.. 3 3... 3 . 3. 3.. 5.339.; P: :89: .3
3. 3.. 3. 3. 3. 3. 3. 3. 3. 3.. 3 3. 3 3.. 3. 09.2358. 3.38: .3
3. 3 3.. 3. 3.. 3.. 3. 3. 3.. 3. 3 . 3. 3 . 3. 3.. 53232.. :33 .nn
3. .N. 3.. 3.. 3. 3. 3. 3. 3. 3. 3 . 3. 3 . 3. 3.. soc-:35... .333 .3
3. 3. 3. 3. 3.. 3.. 3. 3. 3.. 3. 3 . 3.. 3 . 3. 3.. 3353. ..353 .3
3. 3.. —~. —~. 3. 94... 3.. 3. 3. 3.. 3 3.. 3 3. 3. 38. 3.53 .3
3.3 .333 .393 323. 8.2.33

3 jllfi a 3 n a ”~13 5'3 2 a t .uflll

 

 

136

 

 

 

 

e—.. .3.
co. ~n.
se.- we..
we. “9..
.. ee..

me. a...
.m. an..
ea. ~3..
we.. N...
we.. en..
me.. $5..
. s—..

8.
me.
e..

~e..
qe..
ee..
~e..

 

fl neflflllfillflfls 3

On um um an an «a an MN

oe.. we..
ee.. 33.
03.. me.
ee. 03.
me.. 33..
——.. Ne.
we. e3.
~—. se..
se.. no.
we.

nn..
«a..

me.
ee.
0...
nu..
ee.
ee.

8.
~¢..
«3.

1e..
se..
8.

0e..
ee.
nN.
«3..
“3..
se..

8.
up.
up.

n—..
~e..
ne..
mp.
me.
ee.

233.: Ezra one.» 3: .3
now .me

02 .3

83...... 2.19.8.8

8.338. t. :89: .3
8:225... I82... .3
8:389... 6.1....» .3
oo:oco~:.at 09.nuao .ee

3.)...» .353 .on

oucccouc.ox loo. o:.c.e .en

co.u-I;o.c. .o coo.a yea so. no..cul.e ace—on

nn..
pn..
N'.

on..

e¢..
mm.
on..
sn..
3..

53928.... P: 5...: .3
3:235... 89::- .3
8:38.... 5.85» .3

oucocouc.ot ~9.ou:o .en
8.28 .858 .3.
noon ec.¢.e .Nn

...-u hogan-cu. sauce ec.~u.. co.ocnu.e

 

oeauuumllllu

 

 

"7'13 @fl’flifliflflmfllflflmfliflr