RELAUONSHIP OF BlRTH ORDER,
SIBUNG SPACING, AND FAMiLY SIZE
T0 DRINKENG PROBLEMS AND
ORAL BEHAVSOR

Thesis for the Degree of M. A.
MICHIGAN STATE UNIVERSiTY.
HOLLY VAN HORN
1969

lfl

/

8514”]:

M

M

//

' tiff/1W! / MI?

 

W L—

312

, ' -"r ’3 a:
L ' “99¢ Ii} i

was

ABSTRACT

RELATIONSHIP OF BIRTH ORDER, SIBLING
SPACING, AND FAMILY SIZE TO
DRINKING PROBLEMS AND
ORAL BEHAVIOR
By

Holly Van Horn

The present study was designed to examine the re-
lationships of birth order, spacing between siblings,
and family size with problem drinking and oral behavior.
It was hypothesized that later birth positions, close
spacing and increased family size would be related to
greater problem drinking and oral behavior because of
more oral and attentional frustration on the part of
the parents.

Quantity—frequency scores of alcoholic intake,
scores from the Park Problem Drinking Scale, and cigarette
smoking information was obtained by questionnaire from
the subjects, 10“ boys, aged 16 to 18, as well as birth
order, sibling space, and family size. Statistical
analysis demonstrated that there are no main effect dif-
ferences between birth positions for any measure of orality.
Similarly, family size showed no relation to oral behavior.

The spacing between a subject and his next younger sibling

Holly Van Horn

(post-gap) discriminated significantly for first-borns

on quantity-frequency scores, the Park Scale, and cig-
arette smoking, small gaps showing more of these behaviors
than large gaps. These results were attributed to the
greater anxiety of parents when children are born too
closely after a first-born causing frustration of both
oral and attentional needs. Attention was also directed
to the importance of gap as a partial determinant of

birth position effects in earlier studies, which may
account for some of the discrepancies in the earlier

literature.

RELATIONSHIP OF BIRTH ORDER, SIBLING
SPACING, AND FAMILY SIZE TO
DRINKING PROBLEMS AND

ORAL BEHAVIOR

By

Holly Van Horn

A THESIS

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

MASTER OF ARTS

Department of Psychology

1969

ACKNOWLEDGMENTS

I would like to express my appreciation to Dr.
Robert Zucker, chairman of my thesis committee, for his
guidance in this research, his encouragement and his
assistance in the preparation of this manuscript.
Gratitude is also extended to the other members of my
committee, Dr. Dozier Thornton and Dr. Lucy Ferguson
for their participation and reading of the thesis. I
would also like to thank Dr. Lawrence Messé who provided

statistical consulting.

ii

TABLE OF CONTENTS

ACKNOWLEDGMENTS . . . .
LIST OF TABLES . . . . . . .

LIST OF FIGURES

Chapter
I. INTRODUCTION. . . . . . . .
Introduction.
Literature Review--Sibship
Structure and Drinking Behavior
Literature Review--Determinants of
Personality Differences in
Sibship Structure . . .
Statement of the Problem. . . .
Hypotheses . . . . . . . .
II. METHOD. . . . .
III. RESULTS . . . . . . . .
IV. DISCUSSION . . .
Birth Position, Spacing, and Oral
Behavior . . . . . . . .
Family Size and Oral Behavior . .
V. SUMMARY . . . . .
REFERENCES. . . . . . . . . . . . .
APPENDICES. . . . . . . . . . . .

iii

Page

ii

viii

10
l6
19
21
M2.
A2
50
54
56
62

Appendix

I.

II.

III.

Quantity—Frequency Measure of Alcoholic
intake. . . . . .

Park Problem Drinking Scale.

Frequency of Smoking Behavior

Hartley's Test for Hypotheses 2a,
2b, 3a, and 3b .

Tables for analysis of data in Hypo-
theses 2 and 3 . . . .

iv

Page

63
67
7O

72

7A

Table

10.

11.

LIST OF TABLES

Alcoholic Intake Means and Standard Devia-
tions on Large and Small Post-Gap First
and Middle-Borns-—Raw and Transformed
Data

Analysis of Variance on Transformed Alco—
holic Intake Scores Birth Order and
Post-Gap Analysis

Differences Between Alcoholic Intake Means
Hypothesis la-—Transformed Data

Alcoholic Intake Means and Standard Devia-
tions of Later-Borns and Other Birth
Order and Gap Size Groups--Raw and
Transformed Data

Analyses of Variance Comparing Birth Posi-
tions and Post—Gap Size Groups on
Alcoholic Intake (Transformed Data).

Park Problem Drinking Score Means and
Standard Deviations on Large and Small
Post-Gap First and Middle-Borns

Analysis of Variance on Park Problem Drinking
Scores Birth Order and Post-Gap Analysis

Differences Between Park Problem Drinking
Means for Hypothesis la

Means and Standard Deviations of Post-Gap
Size Groups of Park Problem Drinking
Scores for Hypothesis la. . .

Analysis of Variance Comparing Post-Gap Size
Groups on Park Problem Drinking Scores

Smoking Means and Standard Deviations on
Large and Small Post—Gap First and Middle-
Borns——Raw and Transformed Data .

Page

22

2A

2A

27

28

29

29

32

32

33

35

Table

l2.

13.

14.

15.

16.

IV.I.

Analysis of Variance on Transformed Smoking
Scores Birth Order and Post—Gap Analysis.

Differences Between Smoking Means for Hypo-
thesis lb--Transformed Data.

Smoking Means and Standard Deviations of
Later—Borns and Other Birth Order and
Post—Gap Size Groups-—Raw and Transformed
Data. . . . . . . .

Analyses of Variance Comparing Birth Positions
and Post—Gap Size Groups on Smoking Scores
(Transformed Data) . . . . .

Correlations Between Orality Measures.

Hartley's test for Hypotheses 2a, 2b, 3a,
and 3b. . . . . . . . . . . .

Alcoholic Intake Means and Standard Deviations
on Large and Small Pre-Gap Middle and Last—
Borns——Raw and Transformed Data. .

Analysis of Variance on Transformed Alcoholic
Intake Scores Birth Order and Pre- —Gap
Analysis . . . . . .

Alcoholic Intake Means and Standard Deviations
of First and Later-Borns——Raw and Trans-
formed Data

Park Problem Drinking Score Means and Standard
Deviations on Large and Small Pre-Gap Middle
and Last—Borns . . . . . . . .

Analysis of Variance on Park Problem Drinking
Scores Birth Order and Pre-Gap Analysis

Park Problem Drinking Means and Standard De—
viations for First and Later—Borns.

Smoking Means and Standard Deviations on Large
and Small Pre—Gap Middle and Last—Borns--Raw
and Transformed Data . . . . . .

Analysis of Variance on Transformed Smoking
Scores Birth Order and Pre-Gap Analysis

vi

- Page

35

37

39

A0
AA

73

75

76

76

77

77

78

78

79

Table

v.9.

v.10.

.ll.

.12.

.13.

.1“.

.15.

Smoking Means and Standard Deviations of
First—Borns and Other Birth Positions.

Alcoholic Intake Means and Standard De-
viations on Large and Small Pre-—and
Post-Gaps for Middle—Borns-—Raw and
Transformed Data. . . . .

Analysis of Variance on Transformed Alco—
holic Intake Scores Middle-Born Pre-—
and Post-Gap Analysis . .

Park Problem Drinking Means and Standard
Deviations on Large and Small Pre—-and
Post-Gaps for Middle-Borns.

Analysis of Variance on Park Problem Drink-
ing Scores Middle—Born Pre-—and Post-Gap
Analysis. . . . . . .

Smoking Means and Standard Deviations on
Large and Small Pre—-and Post-Gaps for
Middle-Borns—-Raw and Transformed Data.

Analysis of Variance on Transformed Smoking

Scores Middle-Born Pre--and Post-Gap
Analysis. . . .

vii

Page

79

8O

8O

81

81

82

82

LIST OF FIGURES

Figure

l. Alcoholic Intake of First and Middle-Born
Subjects as Related to Size of Post-
Gap. . . . . . .

2. Park Problem Drinking Scores of First and
Middle— Born Subjects as Related to Size
of Post- -Gap . . . . . . . .

3. Smoking Scores of First and Middle-Born

Subjects as Related to Size of Post-Gap.

viii

Page

25

31

36

CHAPTER I

INTRODUCTION

Introduction

 

While the specific variables contributing to the
genesis of drinking behavior are legion, it is necessary
for a scientific investigation that each variable be
separated and scrutinized individually to arrive at an
understanding of both its unique influence and the
nature of its interaction with other contributory vari-
ables. Once the variables are separated it becomes of
further necessity to discover if these variables are im-
portant in and of themselves or if they are not rather
the surface manifestations of deeper and perhaps more
complex determinants. This particular study takes adoles-
cent drinking problems as the focus of the investigation
and birth order, family size, and length of spacing be—
tween siblings as the antecedent variables to be separated.
Another aspect of the investigation is to discover the
extent to which these variables have some consequence. Is
it because they similarly affect the entire spectrum of

oral behaviors, the drinking of alcohol being one of these,

or are the effects alcohol specific? To provide some
orientation for this study and to give support for both
the theoretical outlines and the experimental design
itself, a review of the literature will follow.

Literature Review-~Sibship
Structure and Drinking Behavior

 

 

Stanley Schachter's (1959) imaginative work on the
relationship between affiliative behavior and birth
order instigated researchers in many areas of psychology
to utilize the birth order variable when studying the
behavior and personality correlates of an experimental
subject. Researchers in the field of alcoholism, having
already used this variable to a certain extent, elaborated
their work and emphasized it more than it had been in the
past. Bakan (19A9), a precursor of Schachter in birth
order research, used a chi—square method of analysis and
found that the population of higher birth rank contributed
more than its expected share to the alcoholic population.
As he said, "Youngest children are more likely to appear
in our alcoholic population than individuals of other
birth ranks." Schachter(l959) interpreted these findings
as being supportive of his conclusions that under condi-
tions of anxiety, later-bornindividuals seek to affiliate
significantly less than first—borns. Alcoholism is gener-
ally considered an asocial mode of behavior (Robins, 1966;
Zucker and Fillmore, 1968; Kalin and Williams, in press),

and the alcoholic individual is often found to do much of

his drinking alone rather than in the company of others.

In other words, when faced with anxiety, a later-born would
be more inclined to choose an unaffiliative coping mech—
anism, e.g., drinking, while a first-born would choose a
more affiliative method, e.g., psychotherapy (Schachter,
1959).

Smart (1963) criticized Bakan for choosing subjects
from a population convicted of crimes involving alcohol.
Perhaps the critical variable in this study was the crim—
inal nature of the subjects rather than the incidence of
drinking problems. In addition, he criticized Schachter
for assuming second and third—borns as early—born without
taking the family size into account. In a family of three,
second and third—borns are later-born. In a study of his
own using alcoholics from alcoholism clinics he corrected
for family size using the Greenwood-Yule method (a method
which transforms the size of a sibship in relation to its
frequency in the normal population as determined by census
values) and found no significant differences between the
expected and the observed occurrence of alcoholism in any
birth rank. He did, however, conclude that it is family
size that is important and that the larger the family the
greater the incidence of alcoholism; that is, those with
one to four children were shown to have significantly fewer
drinking problems than those with five or more children.

Smith (1965) criticized the use of census values

as the norm for transformation because the census only

considers the family living together at the time. It does
not take into account that young people will have more
children living at home while older people have more children
who have already married and left home. In comparing the
alcoholics from Smart's study to medical patients in a
general hospital, there were no significant differences in
representation as a function of the size of sibship. But
here again the question can be raised concerning the ap—
propriateness of the control group. Is a sample of patients
from a general hospital comparable to a sample from the
general population? From this study, can it be assumed
that, because there is no difference between the alcoholic
sample and the medical patient sample, there is also no
difference between alcoholics and the general population

on this dimension?

De Lint (196“), using female alcoholics from a re—
search clinic, demonstrated that overrepresentation of
last—borns in his sample was attributable to an over-
representation of persons not reared by both parents.

That is, later—borns are more likely to have lost one or

both natural parents during the first five years than
early-borns. Perhaps it is this absence that influences
later-borns toward alcoholism, rather than affiliative
tendencies. De Lint's findings are supported by Wahl (1956),
who used male alcoholics from a state hospital. Sibling
position was not related to alcoholism but parental loss

was .

Other studies finding significantly more later-borns
in the population of alcoholics were done by Martenson-
Larsen (1957) and Navratil (1956). The latter felt that
these results occurred because of the constitutional
defects (especially digestive weakness) which more fre-
quently characterize the later-born individual. However,
these findings must be evaluated with the awareness that
Martenson-Larsen did not even consider the case of middle-
borns, while Navratil only studied the last-borns, comparing
them to the expected frequency of last-borns in the general
population. Chen and Cobb (1960), on the other hand, found
that alcoholics with police records or hospitalization were
more likely first-born. Supportive of this is the work of
Moore and Ramseur (1960) which revealed more first-borns
than later—borns when the alcoholics are volunteer patients
rather than those who have been coerced into the hospital.

The confusion of the above results emphasizes the
importance of methodology as a determinant of the specific
conclusions that will be drawn. Subject-pools have enough
diversity to preclude comparison between studies; control
groups are haphazard, and the overall designs are pre-
dominantly loose. Even considering that the structure of
the designs were tightened up and it could be more univer—
sally accepted that 1ater-borns were more likely to become
alcoholics, what has this really meant? The problem still
remains as to why later-borns will be more likely to become

alcoholics. Perhaps with this information more productive

preventive measures can be instituted. And finally, many
of the above disparate results may be caused by the con-
founding of the birth position by such things as family
size and, as will be seen later, the size of the gap be—
tween children. All of these factors must be taken into
consideration and corrected before an effective investi-
gation of the problem can be undertaken.

Suggestions concerning the dynamics behind the
occurrence of drinking problems in later-born individuals
other than the previously mentioned suggestions of de-
creased affiliative needs (Schachter, 1959), increased
family size of the later-born (Smart, 1963), parental
deprivation (De Lint, 196A), and a greater extent of
constitutional defects (Navratil, 1956), have been offered.
Weller (1965) feels that it might be related to his finding
that later-born infants have higher levels of arousal than
first-borns, greater conditioned anxiety, and consequently
greater affinity for any method of anxiety reduction.

Through his study of parental attitudes toward child—
rearing, Henry (1957) discovered that more first-borns are
father disciplined and dominated while later-borns tend to
be disciplined more often by the mother. Because the
mother is so strongly related to need gratification, it
is more difficult for the child to express hostility to
her than it is for the first—born to express hostility
toward the father. This results in more guilt over ag-

gressive tendencies, more turning inward, and consequently

less affiliative behavior on the part of the later-born.
Warren (1966) has offered several suggestions for the
birth order differences. He maintains that perhaps they
are produced by changes in the intrauterine environment
after each birth through length of labor or use of forceps.
The age of the mother would then be crucial both before
and after birth; later-born children naturally having
older mothers. Finally, Dittes (1961) and Staples and
Walters (1961) have separately found that first-borns are
more conforming, leading to the conclusion that the more
conforming an individual the less likely he will parti-
cipate in anti-social activities (e.g., alcoholism).
Sampson (1965) criticized the alcoholic and other
birth order studies for overemphasizing the selection of
the sample at the expense of the data (field designs using
correlational methods) resulting in poor control, and yet
warns against the mistake of overemphasizing the utility
of laboratory methods which limit the number of subjects
and offer little in the way of validity or generalization
to the real environment. In addition, he maintains that
the aforementioned studies are putting too much emphasis
on birth order per se. It is too easy for other variables
to enter in and confound the data. He recommends consid-
ering such variables as the sex of the sibling, the sex
of the subject, age spacing, family size, and cultural

and socioeconomic factors. In other words "order of birth,

in and of itself, is not useful in understanding or ex-
plaining the development of personality and behavior"
(Sampson, 1965)-

In birth order research which is not concerned with
alcoholism there have been steps taken in this direction.
Koch (1955), especially, has been concerned with personality
differences related to these other factors-—sex, spacing,
rank, and culture, and has found many complex interrela-
tionships between them resulting in significant differences
on many personality dimensions. Lasko (195A) also re-
searched this area and concluded that spacing was a more
important influencing factor than sex ratio. And finally,
Miller and Zimbardo (1966), using the size of the space
between siblings as a variable, concluded that large gap
last-borns fell between first-borns and narrow gap last—
borns in affiliative trends, and that the larger the gap,
the more the child resembled a first-born or only child.

It appears therefore that it is important to consider
the size of the gap between siblings, as birth order alone
is not the simple variable it may seem to be on the surface.
Miller and Zimbardo have found that the last—born birth
position is not a homogeneous factor but that wide—gap and
narrow—gap individuals respond differently, at least as
far as affiliative behavior is concerned. If this is true
for last—borns, it certainly is a possibility for the
other birth positions as well. Similarly, if the gap effect

exerts influence on the manisfestation of affiliation it

might also be an integral component in the configuration
of later drinking behavior. At the present, no research
has been carried out concerning the relationship be—
tween gap and drinking problems, a fact which this study
will eliminate. In addition to studying the effect of
gap, it is necessary to consider the interaction of gap
and birth position as the gap may not have a similar
effect on all positions but in some may increase the
behavior and in others decrease it. If the interaction
is not investigated, the same criticism can be directed
to gap research as to birth order research, i.e., that
it treats the phenomenon as much more simplified than it
actually is.

These attempts to analyze exactly which aspect of
the family constellation is important in forming per-
sonality should be extended to research with problem
drinking populations. The spacing between siblings
seems to be an especially important influence affecting
personaltiy and should be investigated along with birth
rank when evaluating the etiology of drinking problems.
It would be valueless, however, to look only at the re—
lationships between birth rank, spacing, and alcoholism
without postulating why these variables would have the
effect they do; that is, it is important to discover what

intervenes between the two observable factors.

10

Literature Review-—Determinants of
Personality Differences in Sibship
Structure

 

 

 

One set of intervening variables that should provide
a very strong influence on this interaction are those re-
lating to child—rearing practices. Sears, Macoby, and
Levin (1957) have found that parents differentially treat
children of the various birth positions and that these
treatments are generally consistent across families. For
example, they have found that delight at the prospect of
the child decreases directly with the number of children;
permissiveness also decreases, and frustration of de-
pendency needs increases. Schachter'(l959) suggests that
the mother feels more confident than with the first—born.
In addition, she does not have as much time to spend with
him and thus does not respond to his needs as quickly or
stay as long. Koch (1955), Bossard and Boll (1956) and
Rosen (1961) all found that parents spoil, show preference,
and talk more with the first—born than the later—born child.
In Bossard and Boll's (1956) work they discussed eight
different character "types" and their most probable sib
position. One type was labelled the social isolate and
was described as secretive, antisocial, stubborn, withdrawn,
"at war with the world," and irresponsible. This type was
most often a later-born child. Coincident with this are
Storer's (1961) results demonstrating that the later-born
child is more anti—social, has a less severe superego,

less impulse control, and less ability to delay gratification

11

leading to generally more aggressive behavior. This
finding is supported by other studies (e.g., P.Sears, 1951;
Dean, 19A7; Gewirtz, 19A8; Patterson and Zeigler, 19Al;
Wile and Davis, 19A1; Koch, 1955; and Haeberle, 1958).
However, Stratton (193A) and Wile and Noetzel (1931) found
no significant differences on the dimension of aggressive
behavior.

When looking at child-rearing practices as possible
reasons for differences of drinking patterns between birth
positions, one seems predominant and that is the approach
to feeding that the mother takes for each child. Sears
e§_al., (1957) found that, first of all, fewer youngest
children were breast—fed and that, no matter what method
of feeding was used--breast or bottle--middle and younger
children were weaned earlier and more severely. There
are several explanations for this phenomenon. One is that
the mother is busier and has less time to spend with the
later-born child. Bottle-feeding saves time as the father
or one of the other children can feed the child while a
mother's presence would naturally be required if she were
breast—feeding. In addition, often the mother's desire for
modesty prevents her from breast-feeding a later-born child
when there are other children present. As weaning time
approaches, the mother may wish to complete the procedure
as quickly as possible so that she will not have to spend
as much time feeding. She will therefore accelerate the
weaning process and be more demanding than with her early—

born children.

12

Although the experimental findings were focused on
birth rank data, Sears et_a1., also recognized the im-
portance of the age gap as an influencing factor. It is
unfortunate that they did not include this factor when
considering the differential child-rearing practices of
parents. It would be interesting to discover whether the
narrowness of the gap influences the breast-feeding or
weaning methods of the mother. Following the reasoning
of Sears that birth order is important because of the de—
creased amount of time the mother can afford to spend
with the child, it would also seem tenable that the nar-
rower the gap the more likely the mother would bottle-
feed and wean more strictly. The experimenters did,
however, discover that the mothers are warmer and have a
more favorable attitude toward the child when there is a
large gap rather than a narrow gap.

There has been some evidence to show that the result
of this early weaning may be frustration of oral needs,
needs which will be satisfied by other methods if the
bottle or breast are denied. The work of Levy (1928) is
an example of this type of research when he revealed that
human infants who are deprived of adequate oral stimulation
by bottle or breast resort significantly more often to
displaced sucking, fingers or other objects, than infants
not so deprived. Freud (1905) went even further in em—
phasizing the importance of oral needs and stated that

the deprivation would determine the entire direction of a

13

child's personality development, those frustrated utilizing
such behaviors as kissing, smoking, eating, and drinking
throughout their lives to make up for this lack.

In extending this frustration of oral needs to the
seeking of gratification through drinking, Zucker (1968a)
suggested that concern with primitive gratification may
have its roots in early frustration experiences—-of either
a specifically oral or of a more general dependency de-
privation type--but that in either case, the problem
drinkers in later life appear to be pursuing modes of sat-
isfaction through excitement and sensation seeking (i.e.,
anti-social behavior). Alcoholic drinking easily fits
into this pattern as one manifestation of the same kind,
that in addition provides oral gratification. In this
context, Zucker and Fillmore (1968) found that problem
drinking adolescents had greater amounts of TAT fantasy
concerned with immediate bodily experience——specifica11y,
more physical aggression and more oral ingestion fantasies.

In the same manner, Rorschach cards elicited more
oral responses by alcoholics than controls in a study by
Bertrand and Masling (1969) as well as in a study by
Wiener (1956). Related to this is the work of Masling,
Weiss, and Rothschild (1968) who showed that later-born
individuals responded with more oral images to Rorschach
cards than first and only—borns, perhaps suggesting an ex-
planation for the finding that later-borns are more highly

represented in alcoholic populations.

1A

Other investigators also offer evidence that orality
and drinking problems are concomitant behaviors. Matarazzo
and Saslow (1960) demonstrated that individuals with a
higher alcoholic intake likewise drink more coffee and
smoke more cigarettes than those individuals with a lower
alcoholic intake. Similar findings have been presented

by McArthur, Waldron, and Dickinson (1958) and Heath (1958).

Statement of the Problem

 

The purpose of this study is to discover what the
relationships are between birth position, size of age gap
between children, drinking behavior, and concern with oral
gratification in a sample of late adolescents. It is an
attempt to demonstrate that if in fact later-born and nar—
row—gap individuals are more likely to become problem
drinkers, they will also exhibit a greater amount of oral
behavior in areas other than problem drinking.

If confirmed, this might add support to the theory
that oral deprivation at the infant level is a contributing
cause to later problem drinking. An alternative explanation
is that the narrow age gap is not only important because of
its effects on the feeding practices, but also because it
increases the environmental stress by decreasing the amount
of attention the child is able to receive from the parents.
If this is true, then it would be expected that a narrow
pre—gap (between the subject and older siblings) rather
than only a narrow post-gap (i.e., between the subject

and a sibling who arrives before weaning is completed)

15

would have this effect of increasing drinking problems.
Likewise, a large number of siblings in the family would
decrease the amount of attention the child is able to re—
ceive and should also serve to result in a similar pattern
of drinking behavior. In other words, if a gap effect
were found in the data this would not entirely prove that
it is oral frustration which is causing the result, as

the instigation may come from other stress factors. But
if the post—gap effect is shown to be more influential
than the pre-gap effect this would place more evidence in
the direction of a feeding practices causation than if
both types of gap are equally effective. Optimally, it
would be necessary to actually go back and observe whether
those individuals with later drinking problems have ac-
tually been orally frustrated.

Finally, the use of late adolescents as subjects
rather than alcoholics is important for two reasons. First
of all, the majority of studies employing alcoholics uti-
lize subjects who are by no means representative. Subjects
are taken from clinics, prisons, or social agencies,
leaving the population of less—visible alcoholics, those
not in trouble with the law or seeking help, ignored. It
is unknown whether extrapolation to more representative
populations would produce similar findings. In addition,
by studying the motivations for adolescent drinking, it

may be possible to learn which variables are contributing

16

to subsequent problems before they emerge in the form of
confirmed alcoholism, thereby providing information for

possible prevention (cf. Zucker, 1968b; Zucker and Fillmore,

1968).

Hypotheses

 

la. Post-spacing--First and middle—born adolescents
with a small age—space (0-24 months) between self and next
younger sibling will exhibit greater drinking problems
than first and middle—born adolescents with a large age-
space (more than 2A months) between self and next younger
sibling. Because last—borns have an infinitely large
post—space they are a special case and should be compared
to the other groups in a subsequent analysis.

1b. Post-spacing—-First and middle-born adolescents
with a small age-space (0-24 months) between self and next
younger sibling will exhibit greater orality other than
drinking problems, when compared to first and middle-born
adolescents with a large age—space (more than 2A months)
between self and next younger sibling. Because last-borns
have an infinitely large post-space they are a special case
and should be compared to the other groups in a subsequent
analysis.

2a. Pre—spacing——Middle and last—born adolescents
with a small age—space (0-24 months) between self and next
glggr sibling will exhibit greater drinking problems than

middle and last-born adolescents with a large age-space

17

(more than 2A months) between self and next older sibling.
Because first-borns have an infinitely large pre—space
they are a special case and should be compared to the other
groups in a subsequent analysis.

2b. Pre—spacing--Middle and last-born adolescents
with a small age-space (0—2A months) between self and next
gld§r_sibling will exhibit greater oral behavior other than
drinking problems, when compared to middle and last-born
adolescents with a large age-space (more than 2A months)
between self and next older sibling. Because first—borns
have an infinitely large pre—space they are a special case
and should be compared to the other groups in a subsequent
analysis.

3a. Total space--Middle-born adolescents with a small
(less than 2A months) pre and post space (that is, between
the next older sibling and the next younger sibling, a
total space of A8 months or less) will exhibit greater
drinking problems than middle—born adolescents with a large
pre and post-space (greater than 2A months in each direction,
or a total or more than A8 months).

3b. Total space--Middle-born adolescents with a small
(less than 2A months) pre and post space (that is, between
the next older sibling and the next younger sibling, a total
space of A8 months or less) will exhibit greater oral be—
havior other than drinking problems than middle—born adoles—
cents with a large pre and post space (greater than 2A months

in each direction, or a total of more than A8 months).

l8

Aa. Family size-—As the family size of later—born
adolescents increases, the extent of drinking problems
in adolescents will increase proportionately.

Ab. Family size--As the family size of later-born
adolescents increases, the extent of oral behavior other

than alcoholic drinking will increase proportionatelr.

CHAPTER II

METHOD

Subjects were 10A boys, juniors and seniors in high
school, aged 16-18, members of the cross—validation sample
of Zucker's (1968a, 1968b) continuing research on the de-
velopment of problem drinking among adolescents. These
subjects were chosen at random from class lists of the
one public high school in a community slightly under
15,000 people in one of the Middle Atlantic States. The
community is characterized by a fairly wide distribution
of income and educational levels and religious and nation-
ality groups.

Procedure: Subjects (each paid five dollars for
participation) were contacted at home and asked to attend
a two hour questionnaire session in a church educational
building in their community. The study was introduced as
one concerned with teenagers? leisure time activities.
Information was gathered concerning these activities, and
included questions about the consumption of alcoholic

beverages. Questions were also included to determine

19

2O

birth order, age spacing of siblings, family size, and
nonalcoholic oral behaviors.

The alcohol consumption questions, derived from
Cahalan and Cisin (1967), allowed for the computation of
a quantity—frequency measure of self-reported alcoholic
intake (in standard drinks per year). This measure is
the product of number of drinking occasions per year and
average number of standard drinks consumed per occasion.
See Appendix I for the questions from which this informa-
tion was obtained and the quantitative value assigned to
each.

As another measure of problem oral behavior, the
Park (1962) Problem Drinking Scale items were also ob—
tained. These items were designed to measure such things
as antisocial behavior associated with drinking, the im-
pairment of social relations because of drinking, drinking
to excess, the occurrence of blackouts, benders, etc. The
instrument, together with scoring directions for the items,
can be found in Appendix II.

Nonalcoholic oral intake was assessed via a measure
of smoking amount (cigarettes, pipes, cigars) per week

(see Appendix III).

CHAPTER III

RESULTS

In the main, hypotheses were analyzed by 2 X 2
analyses of variance for unequal cell frequencies (un-
weighted means analysis). The homogeneity of variance
assumption was tested with Hartley's test. In all of
the following analyses, the five only children in the
sample were omitted because of their special characteristics
(i.e., they share characteristics of both first and last-
borns).

For Hypothesis la it was predicted that first and
middle-born adolescents with a small age—space (0-2A months)
between self and next younger sibling will exhibit greater
drinking problems than first and middle—born adolescents
with a large age—space (more than 2A months) between self
and next younger sibling. Means and standard deviations
for quantity—frequency scores for alcoholic intake for
Hypothesis la are given in Table l. Hartley's test for
homogeneity of variance was performed on the data, and

showed significantly different variances (Fm x = 9.73,

a
p < .01). For this reason, and because the data fit the

21

22

TABLE l.-—Alcoholic intake means and standard deviations on
large and small post-gap first and middle—borns——raw and
transformed data.

 

‘2: "

 

 

 

 

 

Raw Transformeda
Classification N

Mean S.D. Mean S.D.

First—born
large post-gap 23 120.75 281.76 13.95 l6.Al
small post—gap 1A 323.50 A91.AA 26.18 2A.71

Middle—born
large post—gap 16 19A.31 A32.15 18.99 20.A6
small post—gap 9 96.06 157.52 1A.Al 13.35

 

a O
Freeman-Tukey square—root transformation.

23

characteristics of a Poisson distribution (one in which
variances increase as means increase and a distribution
which most often occurs when recording number of events
per unit time as in this case) it was decided to perform
a Freeman-Tukey Square-Root Transformation (i.e., x' =
/f + [fir—I; see Mosteller and Bush, 195A). The trans—
formed means and standard deviations are also given in
Table l. The results of the analysis of variance on the
transformed data are shown in Table 2. This shows a
significant interaction (p < .05). Figure 1 illustrates
this relationship; it shows that the greatest contri—
bution to the interaction is from small gap first-borns.
To discover which differences between individual
cell means were significant, a t-test for unweighted means
analysis was applied. Table 3 shows the difference be-
tween all pairs of transformed means and indicates which
differences are significant. The means used to compute
differences represent transformed data. As can be seen
from these data and from Figure l the only significant
difference is between the large and small post-gap for
first-borns. In other words, the post-gap variable appears
to have no effect for the middle—borns and there is no dif-
ference between middle-borns and first-borns as a whole.
The next step was to discover the manner in which
last—borns compare with the other groups. Last-borns are
a special case as they cannot be divided into large and

small post—gap and therefore do not meet the requirements

2A

TABLE 2.--Analysis of variance on transformed alcoholic
intake scores birth order and post—gap analysis.

 

 

Source SS df MS F
Total 2500A.50 62
Post-gap 209.06 1 209.06 0.53
Birth order 161.79 1 161.79 0.A1
Interaction 1592.97 1 1592.97 A.01*
Error 23OA0.68 58 397.25

 

*p < .05, two-tailed test.

TABLE 3.--Differences between alcoholic intake means
hypothesis 1a-—transformed data.

 

Firsts Middles Middles Firsts
large-gap small-gap large—gap small—gap

 

 

Firsts

large—gap -—- 0.A6 5.0A 12.23*

Middles

small-gap --- A.58 11.77

Middles

large gap --- 7.19
*t = 1.85, p < .05, one-tailed test.

Transformed Quantity-Frequency
Means of Alcoholic Intake

27 -
26 ~—
25 ~-
214 ..
23 AP
22 A—
21 q—
20 —-
19 ~-
18 --
17 __

l6 -*-

 

l3

25

First-borns

Last—borns

“—TZ______
Middle-borns

l I
I I ll

Small Large Infinite

Size of Post—gap

Figure l.—-Alcoholic intake of first and

middle—born subjects as related to size of post-gap.

26

for a 2 X 3 design. This birth position can be considered
either from the perspective of birth order or from the
perspective of gap; i.e., last—borns have an infinitely
large post-gap. The significant interaction on the main
analysis indicates that last—borns should be considered
from both perspectives.

Means and standard deviations for both the gap size
and birth order groups are given in Table A. The results
of Hartley's test were significant for both (Fm = 5.02,

ax

p < .01 for the birth order comparison and Fmax = A.93,
p < .01 for the gap size comparison). After the Freeman-
Tukey Square Root Transformation was applied, a one—way
analysis of variance was performed and the results are
given in Table 5. As can be seen, there are no differences
between any of the birth positions or between any of the
post-gap groups. Therefore no evidence is offered that
last-borns differ from any other group on alcoholic intake.
A second set of analyses for Hypothesis la were
carried out using an alternate measure of drinking problems——
i.e., the Park Problem Drinking Scale. For the data from
this scale, means and standard deviations are given in
Table 6. Cell frequencies are identical with those in
Table 1. Hartley's test was performed on the data and
resulted in Fmax = 2.60 (n.s.,) which obviates the need

for any data transformation. Table 7 gives the analysis

of variance results and shows that the post—gap variable

is the only trend effect (p < .10); both the birth order

27

TABLE A.—-Alcoholic intake means and standard deviations
of later-borns and other birth order and gap size groups--

raw and transformed data.

 

 

 

 

Raw Transformeda
Mean S. Mean S.
Birth Order Categorization
first-born 199.60 390.1A 18.58 20.82
middle—born 158.9A 361.50 16.97 16.81
last—born 322.87 809.65 21.62 28.7A
Gap Size Categorization
small post-gap 23A.76 A11.02 21.61 21.76
large post—gap 151.72 35A.8A l6.Al l8.A3
last—born (infinite) 322.87 809.65 21.62 28.7A

 

aFreeman—Tukey square root transformation.

28

TABLE 5.-—Analyses of variance comparing birth positions
and post-gap size groups on alcoholic intake (trans—

formed data).

 

 

 

 

 

Source ss df MS F 1
5.
Birth Position g
Total 62590.27 103
Between 599.50 2 299.75 0.A9
Within 61990.77 101 613.77
Gap Size
Total 757.0A 103
Between 27.6A 2 13.82 1.91
Within 729.AO 101 7.22

 

29

TABLE 6.—-Park problem drinking score means and standard
deviations on large and small post-gap first and middle-
borns.

 

 

Classification Mean Standard Deviation
First-born
large post-gap 2.35 2.28
small post-gap A.57 3.0A

Middle—born
large post-gap 3.19 2.2A

small post-gap 3.00 1.88

 

TABLE 7.--Analysis of variance on Park problem drinking
scores birth order and post-gap analysis.

 

 

Source SS df MS F
Total 38A.61 61
Post-gap 17.56 1 17.56 2.82*
Birth-order A.71 1 A.7l 0.76
Interaction 1.25 1 1.25 0.20
Error 361.09 58 6.22

 

*p < .10, two—tailed test.

3O

factor and the interaction were not significant. Exami-
nation of the means (See Table 6) and Figure 2 show that
the small gap first-borns have higher problem drinking
scores than do the corresponding large gap first—borns.

Because the post-gap effect was significant, sub-
analyses were performed to discover if the effect was
similar for both first-borns and middle—borns. A t—test
for unweighted means analysis was performed. Table 8
shows that only among first—borns is this effect signif—
icant.

Since the only effect that was significant for the
problem drinking data was the post—gap effect, when con-
sidering the last—borns, it is only necessary to compare
them to the large—gap and small-gap groups, rather than
to the other birth positions. The appropriate means and
standard deviations for this comparison are given in
Table 9. Hartley's test yielded Fmax = 1.52 (n.s.) which
does not substantiate the use of a transformation. There—
fore a one-way analysis of variance was performed on
untransformed data; the results are presented in Table 10,
demonstrating no differences between the gap sizes on the
problem drinking dimension. Although the specific re-
sults very somewhat, the general pattern is the same,
demonstrating an interaction between birth position and
post-gap spacing--with this effect being primarily at—

tributable to the difference among first—borns. The small

Mean Scores from Park Scale

31

 

5.0 ——
“—72L—————- First-borns

A.0 ~r

:2 Middle—borns

3.0 Jg
Last—borns

2.0 ~—

1.0 “r-

1 1 l

 

 

I r 1

Small Large Infinite

Post-gap Size

Figure 2.——Park problem drinking scores
of first and middle—born subjects as related to
size of post—gap.

32

TABLE 8.——Differences between Park problem drinking means
for hypothesis 1a.

 

 

Classification Difference t
First-borns: r1
|
Small-gap vs. large—gap 2.22 2.70* i

Middle—borns:

Large-gap vs. small—gap 0.19 0.19

 

*p < .01, one—tailed test.

TABLE 9.—-Means and standard deviations of post—gap size
groups of Park problem drinking scores for hypothesis la.

 

 

Classification Mean Standard Deviation
Large post-gap 2.7A 2.21
Small post—gap 3.96 2.76

Last-born 3.70 2.86

 

33

TABLE 10.-—Ana1ysis of variance comparing post—gap size

groups on Park problem drinking scores.

 

 

Source SS df MS F
Total 757.0A 103
Between 27.6A 2 13.82 1.91
Within 729.A0 101 7.22

 

Ir.

~"“ X ‘i‘t—H I.

3A

gap group shows both greater alcoholic intake and greater
problem drinking than the large gap group.

Hypothesis lb was similar to la—-i.e., the expec-
tation was that post-gap and birth order effects would be
found with other measures of orality than those concerned
with alcohol consumption. Means and standard deviations
for the smoking data are given in Table 11. Hartley's
test gave the following: F = 15.6A (p < .01) re—

max

futing the hypotheses of homogeneity of variance and

‘ . < --a- ch“— -J-r
fl

supporting the use of the Freeman-Tukey Square Root
Transformation. After transformation, a 2 X 2 analysis 1
of variance with unequal cell size was performed with
the results presented in Table 12. This shows that, as
with the alcoholic intake data, the only trend effect is
the interaction. Figure 3 demonstrates this effect more
clearly and it can be seen that while the small post—gap
is associated with increased smoking for the first—borns,
it is associated with decreased smoking for the middle—
borns.
To discover which differences between means are
resulting in the significant interaction, a t—test,
similar to the one used in Hypothesis la, was applied.
Table 13 offers the difference between all pairs of means
and indicates which are significant. The means used to

compute differences represent transformed data. Similar

35

TABLE ll.——Smoking means and standard deviations on large
and small post—gap first and middle-borns--raw and trans-

formed data.

 

 

 

 

Raw Transformeda
Classification N

Mean S.D. Mean S.D.

First-born
large post—gap 23 6.83 20.A9 3.15 A.32
small post-gap 1A 5A.29 80.77 9.32 ll.A7

Middle-born
large post—gap 16 22.98 A9.55 5.39 8.01
small post-gap 9 12.78 20.A3 A.70 5.51

 

aFreeman—Tukey square-root transformation.

TABLE l2.-—Ana1ysis of
scores birth order and

variance on transformed smoking
post-gap analysis.

 

 

Source SS df MS F
Total 3891.29 61
Post—gap 107.10 1 107.10 1.7A
Birth-order 20.28 1 20.28 0.33
Interaction 192.13 1 192.13 3.12*
Error 3571.78 58 61.58

 

*p < .10, two—tailed test.

Transformed Smoking Means

10

36

 

 

 

I-
._A—
; First—borns
,L
1
!.
-—i-— . j .
Last-borns g.
+— . I
‘_:z .
__ Middle-borns
l 1 l
I l l 3
Small Large Infinite

Post—gap Size

Figure 3.—-Smoking scores of first and
middle—born subjects as related to size of
post—gap.

 

37

TABLE l3.--Differences between smoking means for hypothesis
lb--transformed data.

 

Firsts Middles Middles Firsts
large-gap small-gap large-gap small-gap

;-- rhu- Lag—t: Apr

 

an‘

 

Firsts TI
large-gap --— 1.55 2.2A 6.17* :j
Middles
small-gap --- 0.69 A.62
Middles
large-gap --- 3-93

*t = 2.3A, p < .025, one—tailed test.

38

to the alcoholic data the one significant difference was
that between large and small gap first—borns, with greater
smoking for the small—gap group.

Because of the significant interaction, it is neces-
sary to consider last—borns as both a birth order and a
post—gap variable. Means and standard deviations for both

the gap size and birth order groups are given in Table 1A.

The results of Hartley's test were significant for both A1
(Fmax = 3.03, p < .01 for the birth order comparison and g
Fmax = 3.71, p < .01 for the gap size comparison). After

the Freeman—Tukey Square Root Transformation was applied, 21

a one-way analysis of variance was performed and the re—
sults are given in Table 15. As can be seen, there are
no differences between any of the birth positions or be—
tween any of the post—gap groups. Therefore, as with the
alcohol variables, no evidence is found that last—borns
differ from any other group on smoking behavior.

For the remainder of the hypotheses no significant
effect was found for any analysis. The structure was
identical to that in Hypothesis 1; that is, if the test
for heterogeneity of variance was significant a Freeman—
Tukey Square Root Transformation was performed. As can be
seen in Appendix IV this was necessary for the alcoholic
intake data of Hypotheses 2a and 3a and the smoking data
of Hypotheses 2b and 3b. Then a 2 X 2 analysis of vari—
ance for unequal cell frequencies (unweighted means

analysis) was carried out and because no effect even

39

TABLE 1A.——Smoking means and standard deviations of later-
borns and other birth order and post—gap size groups—-raw

and transformed data.

 

 

 

 

Transformeda
Mean S.D. Mean S.D.
Birth order
categorization
First-born 25.A8 62.23 5.A8 8.39
Middle—born 19.31 Al.78 5.1A 7.22
Last-born 39.07 72.69 6.89 10.05
Gap size
categorization
Small post-gap 38.0A 67.A2 7.51 9.86
Large post—gap 13.81 36.71 A.l2 6.28
Last-born (Infinite) 39.07 72.69 6.89 10.05

 

a
Freeman—Tukey square—root transformation.

A0

TABLE 15.——Ana1yses of variance comparing birth positions
and post—gap size groups on smoking scores (transformed

 

 

 

 

 

data).
1...
Source SS df MS F i
E
Birth Position ;
Total 8307.53 103 5
Between 62.05 2 31.02 0.38
Within 82A5.A8 101 31.6A
Gap Size
Total 829A.01 103
Between 220.05 2 110.02 1.38
Within 8073.96 101 79.9A

 

A1

reached the trend level, the first-borns were compared
against the other groups combined for Hypothesis 2. This
was done by use of a t-test, and again no effect was found.
All of these analyses may be found in Tables V.l to V.15 in
Appendix V. As can be seen, no effect reached the trend
level and therefore it is concluded that there is no pre-
gap effect, no birth order effect, no interaction effect
between these two variables and no interaction effect be-
tween pre and post—gap, either for the two measures of
drinking behavior or for the measure of smoking behavior.

Because Hypothesis Aa states that family size and
problem drinking should increase concomitantly, a cor-
relation would be the most appropriate test of this inter-
action. Pearson product moment correlations were computed
and showed no relationship. The r between quantity-fre-
quency alcoholic intake and family size was -.017; the r
between the Park Scores and family size was .07. Neither
of the correlations reach the criterion of .22 required
for significance at the .05 level, showing that there is
no relation between family size and alcoholic drinking
behavior.

Hypothesis Ab similarly states that family size and
nonalcoholic orality should increase concomitantly. The
Pearson product moment correlation between cigarette smoking
and family size was —0.13 (n.s.) thus also showing that

this relationship does not in fact hold true.

CHAPTER IV

DISCUSSION

Birth Position, Spacing,
and Oral Behavior

 

 

The results show that the hypotheses concerning the
post-gap effect on drinking problems and orality were
supported in part and that the hypotheses concerning the
pre—gap effect and the pre-post interaction were not
supported._ Hypothesis 1a was supported by both the
quantity—frequency scores and the Park Problem Drinking
Scale scores, but the observed results were slightly more
complicated than had been anticipated. That is, the post-
gap effect only appeared to be related to problem drinking
for the first-born adolescents and not for middle-borns.

In addition, there appeared to be no differences in pro—
blem drinking between any of the birth positions. Similarly
for Hypothesis lb, the post—gap effect was only significant
for first-borns when considering smoking behavior. Middle-
born adolescents were not affected by length of post—gap

and there were no significant differences between any of

the birth positions.

A2

—~.‘fi"
'L‘flu m

 

A3

It appears then that alcoholic intake, scores from
the Park Scale, and amount of smoking are all operating in
a similar fashion. For all three measures significant
differences were observed in similar comparisons between
large and small post—gaps for the first-borns, with the

difference being in the direction of greater alcoholic

intake, problem drinking, and smoking for the small post- L
gap than the large post—gap first-born subjects. Addi— ‘1
tional evidence for the relationships between these three I
different measures of orality can be ascertained by means i
of correlation. A Pearson product moment correlation was I
carried out and the results presented in Table 16. These
demonstrate that though the variables are partially in-
dependent, they all behave in a similar way.
The question still remains-~are these findings re-
levant to other oral behavior as well, e.g., gum—chewing,
kissing, candy sucking, eating, manipulating mouth with
fingers, or even talkativeness? Perhaps alcoholic intake
and cigarette smoking have some other important factor in
common in addition to orality which makes them react simi—
larly. Therefore, to truly conclude that it is the orality
component that is important in each, it is necessary to
carry out research with the other previously mentioned oral
behaviors.

As indicated in the introduction, there already has

been some work in this area. Matarazzo and Saslow (1960)

AA

TABLE l6.-—Corre1ations between orality measures.

 

Criterion measures (N = 99) r

 

Quantity—frequency alcoholic intake
and Park problem drinking scale
scores .AA

Quantity—frequency alcoholic intake
and cigarette smoking .6A

Park problem drinking scale scores
and cigarette smoking .70

 

Note-—These correlations are all significantly dif-
ferent from zero (p < .01).

A5

found that smokers consume more alcohol than non-smokers.
Similar conclusions were reached in the present study even
though the two experimental designs were not identical.
Matarazzo and Saslow compared smokers with non-smokers on
the amount of liquor consumed per week, while this study
did not separate smokers and non—smokers but used the
measure of number of cigarettes per week as related to L1
alcohol consumption per year. As mentioned earlier, this 1‘
correlation was .6A. In the same manner, Matarazzo and i
Saslow found that smokers drink significantly more coffee
than do non-smokers, providing further evidence for the
concurrency of oral behaviors.

Although not concerned with oral behavior as such
but with fantasies of an oral nature, Bertrand and Masling
(1969) found that alcoholics gave more food responses to
the Rorschach cards than a control group, and that, ex-
cluding alcohol responses, alcoholics gave more oral
dependent responses (e.g., gifts, food organs). Again,
the theory that drinking problems and orality are posi-
tively related is supported.

Post-gap spacing and its relationship with orality
has already been discussed in a previous section (i.e.,
that oral frustration is more probable when another sibling
is born soon after the subject). The fact that the post-
gap has been shown in this study to be a more important
variable than birth position is supportive of Sampson's

(1965) position, namely that the order of birth cannot be

A6

isolated from other important influences. In this context,
the present results are in accord with Miller and Zimbardo's
(1966) study. While the specific results differ, both
studies show the importance of spacing between siblings as
an effect that either interacts with, or is more powerful
than birth position itself.

While the post-gap effect was found to be important,
this study could not confirm the findings of any of the

research concluding that birth order in and of itself was

_ , -, . “mum—r—
r.
. __ _. _ .__-

a distinguishing criteria for individuals with drinking 1;
problems (Bakan, 19A9; De Lint, 196A; Martenson—Larsen, Li
1957; Navratil, 1956; Chen and Cobb, 1960; and Moore and

Ramseur, 1960). The results of these studies not only

disagree with the present study, but also with each other.

This is partly because each study has used a different sub-

ject group, (e.g., criminals, in Bakan, clinic patients,

in Moore and Ramseur, men, in Navratil, women, in De Lint)

and partly because they have divided these subjects into

varying groups (e.g., firsts vs. later-borns, in Moore and

Ramseur, firsts vs. middles vs. lasts, in Bakan, last-borns

vs. the expected frequency of last-borns in Navratil). In

addition to these, however, there is also the fact that a

particular configuration may be confounded by a peculiar

gap characteristic. For example, if the first—borns con—

tained an unusual number of small gap subjects, then a

first—born effect would be obtained. In the case of the

A7

last-born the large variability might under certain con—
ditions produce a last-born effect.
In addition to the fact that there is no pure birth
order effect, the length of spacing between an individual
and his siblings also has no effect unless the individual
is first—born in which case the length of the post—gap is
significant, shorter post—gaps being related to greater l
smoking and drinking problems than large post-gaps. This I
may be the result of either differential weaning or at-
tentional factors, or both. It then becomes necessary to
understand why these factors would be effective for first- L‘é'i
borns and not middle—borns. It may be that the arrival
soon after a first-born of another infant is perceived
as much more threatening and anxiety—arousing to the parents
than the arrival of a baby soon after a middle-born. With
a middle-born the mother is already more comfortable in
the mothering situation and feels more confident in her
abilities to carry out her duties. The decrease in par-
ental anxiety after each child may become less significant
as more children are added to the family since the greatest
decrease occurs after the birth of the first child. In the
same way, the reduction in attention may not be a linear
function of number of children but rather a negatively de—
creasing function. That is, the birth of the second-born
decreases the mother's attention of the first—born more
than the third—born decreases her attention of the second-

born, and so on, down to the last—born. This may be

A8

primarily a contrast effect--the first—born having had

more attention and nurturance in the first place, suffers
greater relative deprivation when a sibling is born soon
afterward than if a sibling is born soon after a later-born.
There may be a point beyond which the birth of another
child has negligible effects on the amount of attention

the other children receive. The present data suggest that
it is only with the first-born that the spacing is of

critical importance.

. .ah‘c'fl—‘nnvzr-V'
,.
_. _ _ . . ._ __fl

Other data support this line of reasoning. The birth
of the first child is greeted with both more anxiety “I
(Schachter, 1959) and more delight (Sears, Macoby, and
Levin, 1957) and warmth (Koch, 1955; Bossard and Boll,
1956; and Rosen, 1961) than later-borns. The delight and
warmth may be enough to make up for the anxiety, but only
if there is enough time to learn to cope with the new
parental role before another child is born. If the mother
has been unable to meet the infant's needs, the ensuing
deprivation may be great enough that the child spends the
remainder of his life trying to make up for this lack,
displacing his "approach" responses to objects which sat—
isfy his needs without also evoking "avoidance" responses.
If this deprivation took its primary form in the frus—
tration of oral needs it may be this area that the indi-
vidual seeks when he perceives stress. Once the mother

has had one child she may be skillful enough at meeting

A9

an infant's needs, both oral and otherwise, that she is
less likely to cause a serious deprivation for her later-
born children even when they are closely spaced.
Some nonsignificant differences are sufficiently in-
teresting also to deserve some comment. Last—borns show
drinking means that are greater than first-borns, evidenced
by quantity—frequency scores of 322.87 and 199.60 respect— L
1

ively. The fact that this is not significant (t = 0.89)

._‘—._ __._,___ _ n . .I_
.

supports the theory that the post—gap——birth order inter-

action is more important than birth order itself, but the

‘1‘- . unq——<..

direction of this difference suggests there may be a

tendency for this frustration of needs to be slightly greater
for last—borns. In addition, when first—borns are broken
down according to gap, the small gap (mean = 323.50)
adolescents show drinking behavior much more similar to
last-borns than do first—borns with a large post-gap (mean =
120.75). Even so, the difference between large and small gap
first-borns is significant (t = 1.85; p < .05, one-tailed
test), while the difference between large-gap first—borns

and last—borns is not (t = 1.A5). The reason for the latter
results can be found in the considerally greater variance

for last—borns compared to that of small-gap first—borns,

the ratio being about 2.7 : 1. This occurrence makes it

much more difficult to achieve a significant effect. In
fact, in all analyses in which last-borns were included,

the variance is larger for this group than any other group,

whether by gap or birth position. Given this greater

50

variability, it is conceivable that in some samples a last—
born effect would be observed. It remains for further
research to isolate what special factors may account for
this greater variance of the last—born group.

When the middle birth positions are considered, the
data neither demonstrate a gap effect (either pre or post)
nor a birth position effect. With regard to oral behavior,
the present data suggest that it makes little difference
how soon other children are born after the first sibling

comes along.

Family Size and Oral Behavior

 

Neither Hypothesis Aa or Ab (the family size hypo—
theses) were supported by this study. There was no
evidence that there is any relationship between family
size and oral behavior——alcoholic or otherwise. These
findings substantiate those of Smith (1965) and contradict
those of Smart (1963). Smith had originally criticized
Smart for overrepresenting the family size of alcoholics
by comparing them to census values which publish smaller
families than in fact exist in the general population.
This study has supported Smith's reasoning and demon—
strates that there is in fact no relation between problem
drinking and family size.

A positive relationship between family size and
orality was predicted in the present study——as family

size increases, drinking and smoking behavior should

51

increase at the same rate. The reasoning behind this was

that oral frustration is more likely to occur when the

mother has many other children to care for. The results,

however, do not support this reasoning. As mentioned in

the discussion of gap size and birth order, the most crucial

time is after the birth of the first child. As long as the

parents have enough time subsequent to the first birth to L
become comfortable in the parental role and in the satis- ii
faction of needs, it becomes unimportant how many other
children are born or how closely they are spaced. While
there may be other stress factors at work in large, closely-
spaced families, they are not of the type to be influential
toward increased orality.

In addition to not supporting the prediction of this
study, the results also do not support the findings of
Barry, Barry, and Blane (1969). In the Barry et_al., study,
they found that delinquents from small families engaged in
more marked and episodic drinking than delinquents from
large families. Conversely in the present study, there is
no indication that there is a negative relationship between
family size and alcoholic involvement. This discrepancy of
findings may be attributable to the fact that in addition to
alcoholic involvement, Barry §t_al.,'s subjects were also
characterized by anti—social behavior, a variable that the
current study did not consider. In their investigation,

the experimenters derived four levels of alcohol involvement

52

ranging from minimal to marked. Small families were de-
fined as two to four children (there were no only children)
while large families were defined as five or more children.
Of the small families 23 delinquents were characterized at
the two greater levels of involvement as opposed to 12
delinquents in the lesser levels of involvement. The sit-
uation was reversed for the large families—-l2 delinquents
were designated as having greater alcohol involvement while
17 were designated as having lesser involvement. It is
unknown if these differences are significant but the trend
is in the direction of adolescents from smaller families
having greater alcoholic involvement than adolescents from
large families. Barry gt_al., were more concerned with
position of birth, however, and emphasized the fact that
in large families these delinquents came more often from
the first-half of the sibship while in small families they
were more often in the second half. The present study did
not investigate the interaction between family size and
position of birth so there is neither confirmation nor
contradiction offered for these findings.

In conclusion, it is evident that further research
is indicated to fully understand the dynamics underlying
drinking behavior of adolescents. This study did reveal,
however, that the length of the space between a first—born
and his next younger sibling is a significant contribution

to his later drinking habits. The shorter age—space may

53

lead to an increase in pathological orality because of the
increased frustration of oral needs and/or environmental
stress caused by the decrease in attention on the part of

the mother.

CHAPTER V

SUMMARY

This study was concerned with the relationships of ‘
birth order, age—spacing between siblings, and family 7
size to alcoholic drinking problems and smoking behavior. 1.
Subjects were 10A boys, aged 16 to 18, members of the
cross-validation sample of Zucker's (1968a, 1968b) con—
tinuing research on the development of problem drinking
among adolescents. Quantity—frequency scores of al-
coholic intake, scores from the Park Problem Drinking
Scale, and cigarette smoking information, were obtained
from questionnaires. Birth order, sibling space, and
family size were collected from this same questionnaire.
Subjects were divided into groups——first—borns into large
(more than 2A months) and small (0—2A months) post—gap
(space between self and next younger sibling), last-borns
into large (more than 2A months) and small (0-2A months)
pre-gap (space between self and next older sibling), and
middle—borns into both large and small post-gap, as well

as large and small pre—gap groups. A 2 X 2 analysis of

5A

55

variance (unweighted means analysis) was performed to dis-
cover significant effects caused by birth order, the two
kinds of gaps, and family size on oral behavior. The re-
sults demonstrated that there are no main effect differences
between birth positions for any measure of orality. Simi-
larly, family size showed no relation to oral behavior.

The post-gap effect discriminated significantly for first-
borns on quantity-frequency scores, the Park Scale, and
cigarette smoking, small gaps showing more of these behaviors
than large gaps. These results were attributed to the
greater anxiety of parents when children are born too close—
ly after a first-born causing frustration of both oral and
attentional needs. Attention was also directed to the im—
portance of gap as a partial determinant of birth position
effects in earlier studies, which may account for some of

the discrepancies in the earlier literature.

REFERENCES

REFERENCES*

Bakan, D. The relationship between alcoholism and birth
rank. Quarterly Journal of Studies on Alcohol, 19A9,
lg, A3A-AAO.

 

Barry, H., Barry, H., & Blane, H. T. Birth order of de—
linquent boys with alcohol involvement. Quarterly
Journal 9£_Studies on Alcohol, 1969, 30, A08-Al3.

 

.. g- :urhigdu'u‘lf‘v
J

Bertrand, Sharon, & Masling, J. Oral imagery and alco—
holism. Journal of Abnormal Psychology, 1969, IA,

50-53.

Bossard, J. H. S., & Boll, Eleanor S. The large family
system. Philadelphia, Pa.: University of Pen—
nsylvania Press, 1956.

 

 

Cahalan, D., & Cisin, I. H. American drinking practices:
A national survey 9f behavior and attitudes {£3
lated to alcoholic beverages. Mimeo. Washington,
D. C., Social Research Group, Report No. 3. The

George Washington University, 1967.

 

 

 

 

 

 

Chen, Edith, & Cobb, S. Family structure in relation to
health and disease. Journal g£_Chronic Disorders,
1960, 12, 5AA—567.

 

Dean, Dauphne A. The relation of ordinal position to
personality in young children. Unpublished master's
thesis, State University of Iowa, Iowa City, 19A7.

De Lint, J. E. Alcoholism, birth order, and socializing
agents. Journal of Abnormal and Social Psychology,

196A. _6__9_, Air-1158‘.“

 

 

*CAAAL notations in the following list refer to ab—
stracts of the Classified Abstract Archive of the Alcohol
Literature. Where CAAAL are given, the author consulted
the abstract rather than the original reference.

57

58

Dittes, J. E. Birth order and vulnerability to differences
in acceptance. American Psychologist, 1961, 16, 358.

 

Freud, S. Three essays on sexuality. Ig_Standard edition
9: the complete works 9f Sigmund Freud. London:
Hogarth Press, 1955.

 

 

 

Gewirtz, J. L. Dependent and aggressive interaction in
young children. Unpublished Doctoral Dissertation,
State University of Iowa, Iowa City, 19A8.

Haeberle, Ann. Interactions of sex, birth order, and de-
pendency with behavior problems and symptoms in
emotionally disturbed pre—school children. Paper
read at Eastern Psychological Association, Phila-
delphia, 1958, (Cited in Maher, B. A. (Ed.).
Progress lg experimental research, Vol. 2. New
York: Academic Press, 1965. Pp. 175-222.)

 

Heath, C. W. Differences between smokers and non—smokers.
AMA Archives 9: Internal Medicine, 1958, 101, 377-
388.

 

Henry, A. F. Sibling structure and perception of dis-
ciplinary roles of parents. Sociometry, 1957, 20,
67—711. ‘—

 

Kalin, R. & Williams, A. F. Personality self—descriptions
of heavy drinking college students. Quarterly
Journal g£_Studies 93 Alcohol, in press.

 

Koch, H. L. The relation of certain family constellation
characteristics and the attitudes of children to-
ward adults. Child Development, 1955, 26, 13—A0.

 

Lasko, Joan K. Parent behavior toward first and second
children. Genetic Psychology Monographs, 195A, A2)
96-137-

 

Levy, D. M. Fingersucking and accessory movements in early
infancy. American Journal 9: Psychiatry, 1928, 1,
881-918.

 

 

Martenson—Larsen, O. The family constellation analysis
and alcoholism. Acta Genetica, 1957, l, AAl—AAA
(CAAAL No. 7675).

 

Masling, J., Weiss, Lillie, & Rothschild, B. Relationship
of oral imagery to yielding behavior and birth order.
Journal of Consulting and Clinical Psychology, 1968,

323 89‘9f

 

59

Matarazzo, J. D., & Saslow, G. Psychological and related
characteristics of smokers and nonsmokers. Psycho—
logical Bulletin, 1960, 51, A93—513.

 

McArthur, C., Waldron, Ellen, & Dickinson, J. The psycho—
logy of smoking. Journal Lf Abnormal and Social
Psychology, 1958,56, 267-275.

 

 

Miller, N., & Zimbardo, P. G. Motives for fear—induced
affiliation: emotional comparison or interpersonal
similarity? Journal pf Personality, 1966, 33, A81-

503.

 

Moore, R. A., & Ramseur, Frieda. A study of the background
of one hundred hospitalized veterans with alcoholism.
Quarterly Journal Lf Studies Ln Alcohol, 1960, 2;,
51- 67

Mosteller, F., & Bush, R. R. Handbook Lf Social Psychology,
Washington, D. 0.: Addison- -Wesley Publishing Co.
Inc., 195A.

 

 

 

Navratil, L. Alcoholism and birth rank. 0n the problem
of "constitution of the last child." Wien. Klin.
Wschr., 1956, 66, 158-160 (CAAAL, No. 7767).

Park, P. Problem drinking and role deviation. In Pittman,
D. J., & Snyder, C. (Eds.) Society, culture, and
drinking patterns. New York: Wiley, 1962.

 

 

Patterson, R., & Zeigler, T. W. Ordinal position and
schizophrenia. American Journal Lf Psychiatry,
19111, 2:1. i455- i456.

 

 

Robins, L. N. Deviant children grown _p, a sociological
and psychiatric study Lf sociopathic personality.
Baltimore: Williams and Wilkins, 1966.

 

 

 

 

Rosen, B. C. Family structure and achievement motivation.
American Sociological Review, 1961, g6, 57A—585.

 

Sampson, E. E. Study of ordinal position: antecedents
and outcomes. Address at Pitzer College, Claremont,
California. In Maher, B. A. (Ed.). Progress Lg
ggperimental personalipy research, Vol. 2. New
York: Academic Press, 1965. Pp. 175—222.

Sears, Pauline S. Doll play aggression in normal young
children: influences of sex, age, sibling status,
father's absence. Psychological Monographs, 1951,
65, No. 6 (Whole No. 323).

 

60

Sears, R. R., Macoby, Eleanor, E., & Levin, H. Patterns
6£_child rearing. New York: Harper and Row, 1957.

 

Schachter, S. Psychology 6: affiliation. Stanford:
Stanford University Press, 1959.

 

 

Smart, R. G. Alcoholism, birth order, and family size.
Journal 6£_Abnormal and Social Psychology, 1963, 66,
17-23.

Smith, C. M. Family size in alcoholism. Journal 9i
Abnormal Psychology, 1965, 16, 230.

 

Staples, F. R., & Walters, R. H. Anxiety, birth order,
and susceptibility to social influence. Journal 23
Abnormal and Social Psychology, 1961, 6;, 716-719.

Storer, E. N. W. Ordinal position and the Oedipus complex.
Laboratory of Social Relations Harvard University
Bulletin, 1961, Lg, 18-21.

 

 

Stratton, G. M. The relation of emotion to sex, primo-
geniture, and disease. American Journal 6: Psycho—
logx. 193A. 36, 590-595.

 

Straus, R., & Bacon, S. D. Drinking ;p_College. New Haven:
Yale University Press, 1953.

Wahl, C. W. Some antecedent factors in the family histo-
ries of 109 alcoholics. Quarterly Journal 6: Studies
92 Alcohol, 1956, L1, 6A3—65A.

 

Weller, G. M. Arousal effects on the newborn infant of
being first or later-born. American Journal 66
Orthopsychiatry, 1965, 36, 3Al—3A2.

 

Wiener, G. Neurotics' depressives' and alcoholics' oral
Rorschach percepts. Journal 2: Projective Techniques,

19563 2_03 “53-1455 0

Wile, I. S., & David, R. The relation of birth to behavior.
American Journal 66 Orthopsychiatry, 19Al, 1;, 320—

33A.

Wile, I. S., & Noetzel, E. A study of birth order and be-
havior. Journal 6£_Social Psychology, 1931, g, 52—71.

 

 

Zucker, R. A. Genesis of problem drinking and normal drink-
ing in adolescents. Unpublished paper, Michigan State
University, East Lansing, Michigan, 1968a.

61

Zucker, R. A. Sex role identity and drinking behavior
among adolescents, Quarterly Journal 6: Studies 66

Alcohol, 1968, 868—88A (Vol. 29).

Zucker, R. A., & Fillmore, Kaye M. Motivational factors
and problem drinking among adolescents. Presented
at the symposium on "Educational Implications of
Recent Adolescent Drinking Research," at the
International Congress on Alcohol and Alcoholism,
Washington, D. C., September, 1968.

APPENDICES

APPENDIX I

QUANTITY-FREQUENCY MEASURE
OF ALCOHOLIC INTAKE

QUANTITY—FREQUENCY MEASURE

OF ALCOHOLIC INTAKE

Within the last year or so, how often do you usually have
any kind of beverage containing alcohol (like wine, beer,

whiskey, liquor or a mixed drink)?

Once a day or more
About four to six times a week

About one to three times a week.

 

About two or three times a month

About once a 92222.

A few times a ZE§£.-

About once a yggp

Less than once a year.

Only tasted alcoholic beverage

Never tasted alcoholic beverage.

(CHECK ONE ANSWER)

 

Coded Value

(No.

Occasions
Per Year)

(36A)
(260)
(10A)
( 30)
( 12)
( 6)
( l)
(0.5)
( O)
( 0)

When you're drinking, what kind of beverage are you most
likely to have?

 

a.

b.

Beer

Wine or a punch containing wine

Whiskey or Liquor (or mixed drinks such as

cocktails or highballs)

6A

65

d. About half the time Beer, and half the time Wine

e. About half the time Wine, half the time Whiskey
or Liquor

 

f. About half the time Beer, half the time Whiskey
or Liquor

For each of the three beverages the following question was
asked:

If you drink Beer, Wine (or punch containing wine), Whiskey
or Liquor (or mixed drinks) at all, how much do you usually
drink at one time: (CHECK ONE ANSWER)

Coded Value
(No. Standard
Drinks)*

a. Never drink
beer
wine
whiskey or liquor . . . . . . . . . . . . . . ( 0)

b. Less than
one can or bottle
one glass
one drink (or cocktail, or shot or
highball) . . . . . . . . . . . . . . . . . (0.5)

c. About one
glass
can or bottle
drink . . . . . . . . . . . . . . . . . . . . ( 1)

d. Two
cans or bottles
glasses
drinks . . . . . . . . . . . . . . . . . . . ( 2)

 

*
Following Straus and Bacon (1953), the following

conversion scale was used: one 12 oz. can of beer =

one 5 oz. glass of wine = one shot of A5% whiskey. This

scale provides essential equivalence between drinks,

in terms of the amount of alcoholic content they possess.

Three
glasses
cans or bottles

Four to six
glasses
cans or bottles

Seven to nine
glasses
cans or bottles

More than nine
glasses
cans or bottles

66

3)

5)

8)

10)

APPENDIX II

PARK PROBLEM DRINKING SCALE

PARK PROBLEM DRINKING SCALE

(Subject answers yes or no)

Have you ever felt that you might become dependent on
or addicted to the use of alcoholic beverages?

Have you ever feared the long range consequences of
your own drinking?

Do you like to be one or two drinks ahead without
others knowing it?

Have you ever gone on the water wagon (stopped drinking
completely for a period of time) as the result of self-
decision or the advice of your family or friends?

Have you ever gone on a week—end drinking spree (been
high or tight most of Saturday and Sunday with nothing
worse than a hang-over on Monday)?

Do you usually drink to comply with custom-~that is,
because other people are, or because its polite or
the socially acceptable thing to do?

Score one if any of the following questions (7a-7f) is
answered yes:

7a.

7b.

7c.

7d.

Have you ever been arrested or detained or charged
because of drunken driving or other behaviors result-
ing from drinking?

Have you ever come before school authorities in con-
nection with drinking?

Have you ever had alcohol interfere with your pre-
paration for classes or exams?

Has the use of alcohol ever caused you to miss
appointments?

68

7e.

7f.

69
Has the use of alcohol ever caused you to not be able
to afford other things because of the expense of liquor?

Has the use of alcohol ever caused you to lose close
friends or other friendships?

Score one for each "yes" answer:

8.

10.

11.

12.

13.

Have you ever gotten into a fight or damaged property
while you were drinking?

Have you ever had a drink or two before or instead
of breakfast?

Have you ever drunk so much that you could not remember
afterwards some of the things you had done?

Have you ever had some drinks when you were alone?
Have you ever been drunk?
Drinks one or more times a week and on the average

consumes at least four drinks per occasion. (Determined
from subject's drinking reports.)

APPENDIX III

FREQUENCY OF SMOKING BEHAVIOR

 

 

 

FREQUENCY OF SMOKING BEHAVIOR

How many cigarettes (or pipes, or cigars) do you smoke

now? (CHECK ONLY ONE)

Coded Value

(No.

a. More than a pack (or more than 20) a day
b. About a pack (or 20) a day

0. About 10—19 a day. . . .

d. About 5-9 a day.

e. About l-A a day.

f. Less than one a day, but more than three
times a week

g. One to three times a week.
h. One every few weeks.
i. Once a month or less

j. I don't smoke at all

71

Occasions
Per Week)

(200)
(150)
(100)
(50)
(15)

( 5)
( 2)
(0.5)
(0.2)
( o)

 

APPENDIX IV

HARTLEY'S TEST FOR HYPOTHESES
2a, 2b, 3a, AND 3b

73

TABLE IV.l.--Hartley's test for hypotheses 2a,

and 3b.

2b,

 

Criterion

max

 

2 X 2 Analysis of Variance, Pre—gap

and Birth Order Analysis, Alcoholic

Intake

First—borns vs. Later—borns, Alco-
holic Intake

2 X 2 Analysis of Variance, Pre-gap
and Birth Order Analysis, Park
Scores

First vs. Later-borns, Park Scores

2 X 2 Analysis of Variance, Pre-gap
and Birth Order Analysis, Smoking
Scores

First-borns vs. Later—borns, Smoking
Scores

2 X 2 Analysis of Variance, Pre—gap
and Post-gap for Middle-borns,
Alcoholic Intake

2 X 2 Analysis of Variance, Pre—gap
and Post-gap for Middle-horns,
Park Scores

2 X 2 Analysis of Variance, Pre—gap
and Post—gap for Middle-borns,
Smoking Scores

57.

19.

16A5.

1A.

20

.0A

.00

.15

31

.22

78

.90

93

.01

.01

.01

.01

.05

 

APPENDIX V

TABLES FOR ANALYSIS OF DATA IN
HYPOTHESES 2 AND 3

 

75

TABLE V.1.--Alcoholic intake means and standard deviations
on large and small pre—gap middle and last—borns--raw and

transformed data.

 

 

 

 

a
Classification N Raw Transformed
Mean S.D. Mean S.D.
Middle-born
large pre-gap ll 2A3.Al 507.A7 21.75 22.A2
small pre-gap 1A 92.57 lAA.9A 13.88 13.39
Last-born
large pre-gap 35 288.88 705.62 20.9A 26.81
small pre-gap 8 595.06 1096.2A 29.22 39.09

 

a .
Freeman-Tukey square-root transformatlon.

 

q i L .71)-. “1"“

76

TABLE V.2.——Ana1ysis of variance on transformed alcoholic
intake scores birth order and pre—gap analysis.

 

 

Source SS df MS, . F
Total A692A.93 67
Pre-gap 0.51 l 0.51 0.00
Birth order 672.A2 l 672.A2 0.95
Interaction 830.65 1 830.65 1.17
Error A5A2l.35 6A 709.71

 

 

TABLE V.3.——Alcoholic intake means and standard deviations
of first and later—borns-—raw and transformed data.

 

 

 

 

Classification Raw Transformed
Mean S.D. Mean S.D.
First-horns 199.60 390.1A 18.58 20.82
Later-borns 277.13 680.29 20.59 23.01

 

t = 0.A3 (n.s.) for transformed data.

77

TABLE V.A.--Park problem drinking score means and standard
deviations on large and small pre-gap middle and last—
borns.

 

Classification Mean Standard deviation

 

Middle-born

large pre-gap 3.5A 2.19

small pre—gap 2.78 2.01
Last-born

large pre-gap 3.88 2.88

small pre—gap 2.88 2.80

 

TABLE V.5.—-Analysis of Variance on Park problem drinking
scores birth order and pre—gap analysis.

 

 

Source SS df MS F
Total A6A.09 67
Pre-gap 9.9A 1 9.9A 1.A0
Birth order 0.6A 1 0.6A 0.09
Interaction 0.00 l 0.00 0.00

Error A53.5l 6A 7.09

 

78

TABLE V.6.——Park problem drinking means and standard de-
viations for first and later-borns.

 

 

Classification Mean Standard deviation
First—born 3.25 2.82
Later—born 3.A8 2.63

 

t = 0.A1, n.s.

TABLE V.7.—-Smoking means and standard deviations on large
and small pre-gap middle and last-borns--raw and trans-
formed data.

 

 

 

 

Raw Transformeda
Classification
Mean S.D. Mean. S.D.
Middle-born
large pre—gap 56.77 56.81 7.36 9.00
small pre—gap 8.23 17.A8 3.A0 A.75
Last-born
large pre-gap 38.00 71.67 6.07 10.16
small pre-gap A3.75 76.80 7.36 11.05

 

aFreeman-Tukey square—root transformation.

79

TABLE V.8.--Analysis of variance on transformed smoking
scores birth order and pre-gap analysis.

 

 

Source SS df MS F
Total 5938.8A 67
Pre-gap A2.81 l A2.81 0.A7
Birth order A2.81 l A2.81 0.A7
Interaction 57.58 1 57.58 0.6A
Error 5795.6A 6A 90.56

 

TABLE V.9.—-Smoking means and standard deviations of first—
borns and other birth positions.

 

 

Classification Mean Standard deviation
Middle and last-borns combined 31.80 63.83
First-borns 25.A8 62.23

 

t = 0.51, n.s.

80

TABLE V.10.--Alcoholic intake means and standard deviations
on large and small pre—~and post-gaps for middle-borns-—raw
and transformed data.

 

 

 

 

Raw Transformeda
Classification
Mean S.D. Mean S.D.

Large post-gap

Large pre—gap 7 32A.00 57A.68 25.91 25.03

Small pre-gap 8 6A.62 98.07 12.07 10.71
Small post-gap

Large pre-gap A 21.38 1A.16 8.2A A.39

Small pre—gap 6 129.83 183.76 16.29 15.98

 

aFreeman—Tukey square—root transformation.

TABLE V.ll.--Analysis of variance on transformed alcoholic

intake scores middle-born pre——and post-gap analysis.

 

 

Source SS df MS F
Total 7203.59 2“
Pre-gap A8.86 l A8.86 0.15
Post-gap 263.63 1 263.63 0.82
Interaction 119.89 1 119.89 0.37
Error 6771.21 21 322.AA

 

81

 

TABLE V.12.-—Park problem drinking means and standard
deviations on large and small pre-~and post—gaps for

middle-horns.

 

 

 

 

 

 

Classification Mean Standard deviation
Large post-gap
large pre-gap 3.88 .26 F
small pre-gap 2.50 .03
Small post-gap ,
large pre—gap 2.25 .6“ i
small pre-gap 3.17 .95
TABLE V.13.--Analysis of variance on Park problem drinking
scores middle-born pre--and post-gap analysis.
Source SS df MS F
Total 98.63 2h
Pre—gap 0.29 l 0.29 0.07
Post—gap 1.3“ l 1.3M 0.32
Interaction 7.70 1 7.70 1.81
Error 89.30 21 “.25

 

TABLE V.1“.——Smoking means and standard deviations on large
and small pre—-and post—gaps for middle—borns--raw and

transformed data.

 

 

 

 

Transformeda
Classification
Mean S.D. Mean S.D.
Large post—gap
Large pre—gap 59.69 63.86 7.88 9.89
Small pre—gap 6.28 16.53 2.72 “.35
Small post-gap
Large pre-gap 12.99 21.65 “.30 5.72
Small pre-gap 10.83 18.35 “.90 5.08

h.

 

aFreeman—Tukey square—root transformation.

TABLE V.15.--Ana1ysis of variance on transformed smoking
scores middle—born pre——and post—gap analysis.

 

 

Source SS df MS F
Total 1229.20 2“
Pre-gap 38.01 1 38.01 0.70
Post-gap 5.5“ l 5.5“ 0.10
Interaction “0.75 l “0.75 0.75
Error 11““.90 21 5“.52

 

‘ .. .1. T

 

MICHIGAN STQTE UNIV. LIBRRRIES

31293008523908