Parental favoritism

Material Information

Parental favoritism
Rose, Trina
Publication Date:
Physical Description:
42 leaves : ; 28 cm


Subjects / Keywords:
Parent and child ( lcsh )
Parental acceptance ( lcsh )
Sibling rivalry ( lcsh )
Parent and child ( fast )
Parental acceptance ( fast )
Sibling rivalry ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 40-42).
General Note:
Department of Sociology
Statement of Responsibility:
by Trina Rose.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
57707983 ( OCLC )
LD1190.L66 2004m R67 ( lcc )

Full Text
Trina Rose
B.A., University of Northern Colorado, 1998
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts

by Trina Rose
All rights reserved.

This thesis for the Master of Arts
degree by
Trina Rose
has been approved
Candan Duran-Aydintug
Yili Xu
Susan Allison-Endriss

Rose, Trina (M.A., Sociology)
Parental Favoritism
Thesis directed by Associate Professor, Candan Duran-Aydintug
Previous research has revealed that there are reasons and consequences of parental
favoritism. This line of study has been neglected in academic research until recently.
The present study extends this line of research by examining the Reasons of maternal
favoritism as perceived by the children using a sub-sample from the National Survey
of Children 1976-1987 data set. Results showed that variables such as family
income, race of the mother, and politeness of the child were associated with reports of
parental favoritism. These preliminary results highlight the importance of and need
for further research regarding parental favoritism and parental differential treatment.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Candarf Duran-^dintug

I dedicate this thesis to my family:
I dedicate it to Laura Nay, my mommy, for her unbelievable and undying
emotional, mental, physical, and financial support, who has not only been a mother,
but a mentor, intellectual colleague, and most importantly best friend (keep smiling,
keep shining, knowing you can always count on me).
I dedicate it to Tim Oman, my daddy, for encouraging me to pursue my
educational dreams, and believing I could, even though Im a girl, and for giving
me the work ethic and financial understanding and support to do so (Im still your
I dedicate it to my southern mom, Tawanah Fagan, who could have and
should have been the best biological mother ever, but asked me to be her only baby
instead, for being the rock of stability in my life and giving me a place to always call
home, and for giving me sociology before I even knew what it was (Im still your
Finally, I dedicate it to my brother, Chris Oman, for seeing me through
personal hardships that only a big brother could understand and relate to, for always
keeping me laughing, and for giving me a beautiful niece, Elizabeth Gonzalez, and
nephew, Christian Tabor Oman, whom I love with all my heart.

My thanks to my thesis committee, Candan Duran-Aydintug, Yili Xu, and Susan
Allison-Endriss for their knowledge and support during my graduate program, and to
Richard Anderson, Virginia Fink, and Andrea Harr; for their support also. Finally, I
acknowledge the University of Colorado at Denver, Sociology Department for giving
me the opportunity to pursue my education.

1. INTRODUCTION...............................................1
Consequences of Parental Favoritism.....................3
Reasons for Parental Favoritism........................12
Perceived (vs Actual) Parental Favoritism..............14
Mainstream or Pop-Culture Views on Parental Favoritism.15
Theoretical Perspectives in Existing Literature........16
Socialization Theory...................................18
Demographic Variables...........................21
Other Independent Variables.....................23
Data and Procedure.....................................26

4. RESULTS...................................................29
5. DISCUSSION AND CONCLUSIONS................................31
Discussing the Results.................................31
Suggestions for Future Research........................36
A. Table A1. The wordings, Codings, Means, and Standard
Deviations of Parental Favoritism Variables............37
B. Table A2. Unstandardized and Standardized Coefficients of
Dependent and Independent Variables of Favoritism.........39
REFERENCES ........................................................40

Al. The Wordings, Codings, Means, and Standard Deviations of Parental
Favoritism Variables......................................................37
A2. Unstandardized and Standardized Coefficients of Dependent and
Independent Variables of Favoritism.......................................39

Parental Differential Treatment (PDT), as it is referred to in the literature, is
what I will refer to as parental favoritism. In addition, I will use these two terms
interchangeably throughout this study. My definition of this concept is the act of a
parent favoring one child over another. There can be different combinations of
parental favoritism, which have been given different names. There are congruent
patterns of favoritism, in which parents are more likely to favor or enjoy both
children equally, than to enjoy or favor either one or the other child more. There are
incongruent patterns of favoritism, which is when one parent treats the two children
equally, but the other parent directs more favor to one or the other child. And finally,
there is the complementary pattern of favoritism, in which one parent directs more
positive behavior to one sibling and the other directs more of the behavior to the other
sibling (Volling, 1997).
While there are many different areas to explore regarding parental favoritism,
it has only been studied by a few researchers (Brody, Stoneman, and Burke, 1987;
McHale and Pawletko, 1992; Volling, 1997). Even though parental favoritism dates
back to ancient Greece (Zervas and Sherman, 1987), it seems as if it had been an
unspoken taboo until the late 1980s, and only then researchers directed their

attention to this phenomenon (Brody, Stoneman, and McCoy, 1992). In addition to
this recent attention in academia, pop culture (such as Good Housekeeping) and the
media (such as CBS) have been addressing the issue. In academia, Parental
favoritism has been studied mainly by psychologists and sociologists, and research
has mainly concentrated on either consequences of parental favoritism or reasons for
parental favoritism. I will address these consequences and reasons in the following
literature review. Also, while there are many different theoretical perspectives that
could help explain parental favoritism, I utilize a theoretical framework that has been
ignored by past researchers. Finally, I tested my hypotheses using a sub-sample from
the first wave of the 1976-1987 National Survey of Children (NSC). While there
were many results that were expected and consistent with the literature, there were
also results that were new and interesting, and will definitely add to the base of
knowledge regarding parental favoritism.

Though research in the area of Parental Favoritism and Parental Differential
Treatment is limited, the majority of the existing studies can be categorized into three
broad areas: the consequences of parental favoritism, the reasons for parental
favoritism, and perceived (versus actual) parental favoritism. I have reviewed the
research on each of these areas. Parents who are becoming concerned about their
own parental practices seem to also be asking for help through pop-culture and the
media. This literature review will reflect some of those non-academic studies and
articles also.
Consequences of Parental Favoritism
According to the literature, the exhibit of parental favoritism has a wide
variety of consequences, including lower academic achievement, (Barrett Singer and
Weinstein, 2000), lower self-esteem, (Zervas and Sherman, 1994), inadequate
emotional adjustment, (Volling and Elins, 1998), and sibling rivalry, (Volling, 1997;
and Brody, Stoneman, and McCoy, 1992). Other studies found that parental
favoritism may be normative, and thus have no consequences on a childs behavior
(Crouter, McHale, and Tucker, 1999).

In a quantitative study, Barrett Singer, & Weinstein (2000) used a diverse
sample of male and female Asian American and European American late adolescents
to examine their hypotheses. They speculated that participants would perform better
academically and have more positive self-perceptions if (1) they are more
differentially favored they are by parents (measured by more affection or less
control). Their second hypothesis stated that when there are less discrepancies in
parenting reported, the better participants will perform academically, and the more
positive their self-perceptions.
Barrett Singer and Weinstein (2000) administered questionnaires to 148
undergraduate students enrolled in psychology courses at a large public university.
Half of these students were Asian American (both parents were of Asian descent),
and half of this sample were European American (both parents of European descent).
Sixty-nine participants were the older sibling, and seventy-nine were the younger
sibling. Seventy seven had the same gender sibling, and seventy-one had the opposite
gender sibling.
Barrett Singer and Weinstein (2000) administered the Differential Maternal
Affection and Control and the Differential Paternal Affection and Control portions of
the Sibling Inventory of Differential Experience (SIDE) questionnaire, and a revised
format of the SIDE, the SIDE-R, to the above described sample of students. The
original SIDE is described as a 5-point Likert scale. Pride, enjoyment, sensitivity,
interest in siblings, and favoritism were the five differential affection items, and

parental strictness, punishment, blaming, and disciplining of siblings were the four
differential control items. The SIDE-R was a revised version of the SIDE and has a 4
point Likert scale, and is based on a Differential Teacher Treatment Inventory. The
SIDE-R included (1) separate forms for mother and father, (2) additional questions
regarding interactions with parents were on independent scales, and (3)
questionnaires for self-interactions with mother and father were separated using filler
measures for sibling interactions with mothers and fathers. Results, according to
Barrett Singer and Weinstein, (2000), showed that children who perceived more
parental favoritism, had less achievement scores and more negative self-perceptions.
Zervas and Sherman (1994) also found that self-perception was affected by
parental favoritism. Zervas and Sherman (1994) conducted a study using ninety-one
undergraduate psychology students. Of the ninety-one students, there were sixty-two
females, and twenty-nine males, all of whom were volunteers at the same college, that
is located in the northeast, and is predominantly Catholic. These students were asked
to answer three questionnaires. The Coopersmith (1967, cited in Zervas and
Sherman, 1994) Self-Esteem Inventory (SEI) measured the students self-esteem.
According to Zervas and Sherman (1994), this questionnaire is widely used to
measure self-esteem, showing high reliability and validity. The second questionnaire
measured favoritism. This questionnaire was not based on an existing, validated
questionnaire, because there was not one available. On the last questionnaire,
students were asked to give their demographics. Results showed that parental

favoritism was significantly related to a childs self-esteem. The no-favoritism
group had higher social self-esteem and higher total self-esteem than the non-
favored group.
In addition to self-perception and self-esteem, some research has shown
significant effects of parental favoritism on sibling rivalry. For example, Volling
sampled thirty maritally-intact, Caucasian families with two preschool-age children,
and observed them in their homes on two separate occasions (1997). In this study,
she had both parents introduce toys to children until they felt the children were
comfortably engaged in play on both visits.
After the first home visit, the parents were given questionnaires, including the
SIDE (1985), the Braiker and Kelley (1979) four-factor scale of intimate relations,
and a 25- item questionnaire based on a pre-existing observation of sibling
interactions. The SIDE was chosen to partially replicate a previous study on PDT.
The Braiker and Kelley (1979) questionnaire was given an high internal consistency
ratings for the subscales used.
Volling (1997) was the only person collecting all the narrative observations,
so no formal assessment of narrator reliability was conducted. But all coders were
trained to an 85% reliability, and 10 of the narratives were randomly selected and
rated by two independent coders for purposes of reliability assessment. Volling
defined sibling conflict and play in detail, both of which were rated on a 5-point
scale. In this study, she found that when mothers reported favoring their younger

child, they also reported that both their older and younger children expressed more
hostility toward one another... than in families where parents reported favoring
children equally (p. 234).
Similarly, researchers Brody, Stoneman and McCoy (1992) examined the
different ways fathers and mothers treated different children by video taping
interactions between them. Ninety-eight families participated in this study. They
were separated into four groups based on emotionality ratings: older vs. younger
sibling, or both vs. neither. These ratings were given by the parents. The families
had to have same sex children, where the younger sibling was between the ages of
four to nine years, and the older sibling was between the ages of six to eleven years.
This was to avoid the teenage years. Forty-one pairs of the children were female and
fifty-seven pairs were male. All of the families were Caucasian, middle to upper
middle class families from two southeastern states. Both biological parents had to be
present in the home. Mothers had an average age of thirty-six years and an average
education of fifteen years. Fathers had an average age of thirty-eight years and an
average education of sixteen years. (Brody, Stoneman, and McCoy, 1992)
Two videotaped sessions were held a week apart in the families living room,
and the children and parents were all given the same tasks. Each parent was
videotaped interacting with her/his children, without the other parent there. Then, the
videotapes were evaluated using a behavioral coding system, which coded intervals of
five different behaviors with each child: controlling, positive, responsive, negative,

and verbal behavior. (Brody & Stoneman, 1992) These researchers found that rates
of parental differential behavior were associated with negative sibling behavior (p.
Brody and Stoneman worked also with Burke (1987) on Sibling Rivalry and
PDT, and began by selecting a sample of 40 mothers with same sex children (half
boys, half girls), where the children were separated by at least two years but not more
than three. The older child ranged from seven to nine years, and the younger children
ranged from four and a half to six and a half years, all from Caucasian upper middle-
class families (Brody, Stoneman, and Burke, 1987). Variations of this sample were
videotaped in their living rooms on two separate occasions one week apart, in which
the researcher took a complete observer role. One session videotaped interactions of
the mother-sibling triads, and the other videotaped interactions between the siblings
only. The order of these observations was counterbalanced. During these sessions,
Trouble (a popular board game by Gilbert Industries) and things such as Legos were
used as interaction materials; each activity was given a thirty-minute observation
period. These activities were chosen because according to Brody, Stoneman, and
Burke, (1987), they were common household activities at this time. After a short
introduction activity to get accustomed to the video equipment, the families were
asked to participate in these activities, acting as if the video equipment were not there.
One week following the observation phase, the mothers were visited in their
homes and given a questionnaire (Temperament Assessment Battery TAB), to

assess temperament dimensions in children ages three to nine. Upon completion of
the video observations, two graduate assistants served as coders of the video material.
For the sibling behaviors, verbal, prosocial, and agonistic behaviors were coded. For
the mother behaviors, verbal positive, negative, and managing behaviors were coded.
Interobserver agreement reached eighty-five percent, there were weekly meetings to
assure reliability (Brody, Stoneman, and Burke, 1987). The researchers found that
maternal differential behavior was associated with lower rates of all sibling-to-sibling
agnostic and prosocial behaviors.
In addition to all of the above mentioned consequences, some researchers
showed that parental favoritism had emotional adjustment consequences. Brody,
Stoneman, and McCoy, (1992) found in their study which examined the different
ways fathers and mothers treated different children by video taping interactions
between them, that mothers directed higher rates of controlling, negative,
responsive, and verbal behavior (but fathers directed higher rates of positive and
responsive behavior) to younger siblings when the younger siblings were rated higher
on negative emotionality than their older siblings (p. 8).
Many of the above researchers used parental enjoyment and parental
discipline of children to measure Parental Differential Treatment. Volling and Elins
(1998) addressed this in their study, however, when they found that favoritism of one
child over the other may be reflected in the parents reports of differential enjoyment
of that child, but is not reflected in whether the parent disciplines one child more than

another. They administered four different questionnaires to sixty families, to gain this
information. These families had to be maritally intact, both parents had to agree to
participate, and there had to be at least two children in the family. The questionnaires
were completed by both mothers and fathers. The differential parental treatment
subscale of the Sibling Inventory of Differential Experience (SIDE) questionnaire
was used to measure PDT. Forty-nine items of a modified version of the Sibling
Inventory of Behavior (SIB) was used to evaluate sibling relationships. The Braiker
and Kelley Intimate Relations Scale was used to asses the marital relationship.
Additionally, the Child Behavior Checklist (CBCL) was used to evaluate childrens
emotional outcomes (Volling & Elins, 1998). Results showed that favoritism may be
reflected in the parents reports of differential enjoyment, but it is not reflected in
whether the parent disciplines one child more than another (Volling & Elins, 1998).
As one can see, the literature shows a wide variety of consequences for
parental favoritism. Still, not all researchers are convinced that parental favoritism
has consequences on children. For example, Kowal and Kramer (1997) interviewed
sixty-one Caucasian families. Focus children were between eleven and thirteen years
of age and their siblings were between one and a half and four years older or younger
than the focus children. Their recruitment was through advertisements. If there were
more than two children in the family, they chose to study the two that were closest in
age. All of the families were maritally intact. The dyads of siblings were as follows:
fifteen older sister younger sister, fifteen older sister younger brother, sixteen

older brother younger brother, and fifteen older brother younger sister. (Kowal &
Kramer, 1997) These researchers found that seventy-five percent of the children who
acknowledged that differential treatment was occurring in their homes did not think it
was unfair.
In addition, McHale and Pawletko (1992) studied parental differential
treatment in families with a mentally disabled child, and in families without a
mentally disabled child. In their interviews with sixty-two Caucasian children, where
half of these children had a sibling with mental retardation, and were lower middle-
class, they used interviews, as well as three different pre-existing scales to measure
differential treatment, depression, anxiety, and sibling relationships. Their results
indicated that although some significant correlations between the domains of
differential treatment were found, the lack of overlap was pronounced.
In another study, by Crouter, McHale and Tucker, (1999), one hundred
eighty-seven two-parent families participated in interviews, in which the mother,
father, first-born and second-bom child were interviewed in their home. The sample
included families with approximately equal gender composition of the siblings.
These families were given five preexisting questionnaires (one of which included the
SIDE). These researchers found that parental differential treatment may be

Reasons for Parental Favoritism
Some of the literature is more concerned with why parental favoritism exists,
instead of its outcomes. Just as there are many different reasons that people favor
friends, acquaintances, and co-workers, researchers have found that the reasons
parents favor one child over another vary on a wide continuum. Some of the reasons
given in the research are, stress, (Crouter, McHale, and Tucker, 1999), personal
qualities, (Tucker, McHale, and Crouter, 2003), gender, and family size,
(Konstantareas and Desbois, 2001).
One reason given for parental favoritism is stress. In the study conducted by
Crouter et al., (1999), mentioned in the previous section, it was found that stress
might exacerbate parental differential treatment of siblings.
As might be expected, parents and childrens personal qualities were also
found to be a reason for parental favoritism. Tucker, McHale, and Crouter, (2003),
studied parental differential treatment and sex-typed personal qualities. In their
interviews with one-hundred eighty-eight different families, they conducted separate
interviews with the first and second-bom children, and mothers and fathers. These
families, were all Caucasian, and middle-class from small cities in one northeastern
state. The researchers used pre-existing measurement scales to gather their data: the
Sibling Inventory of Differential Experiences scale, the Antill Trait Questionnaire,
and the Emotionality, Activity, Sociability temperament measure. In this study, the

results indicated that sex-typed personal qualities, such as, leadership,
competitiveness, sensitivity, and kindness were linked to parents demonstration of
differential discipline and affection (Tucker, McHale, and Crouter, 2003).
In addition to stress and personal qualities, gender is cited as a reason for
parental favoritism. In Barrett-Singer and Weinsteins, (2000) study as mentioned in
the previous section, the authors found that reports of differential treatment by parents
toward late adolescents differed according to gender match of siblings.
Konstantareas and Desbois, (2001), also found that gender had an effect on
parental favoritism. Konstantareas and Desbois (2001) presented their vignettes to
fifty-seven preschoolers between the ages of four and six years old from daycares and
preschools in two suburban towns of a large city. These children were from upper-
middle class families, and were assigned to either an intact family group, or a
single parent group. The children were presented the vignettes as if it were a game.
All vignettes began with the same thing: If a little girl/boy was naughty, for
example s/he did not listen to her mom, would it be fair for the mom to ...
(Konstantareas and Desbois, 2001, p. 478). After the child heard the vignette, s/he
was instructed to pull a gauge to one of five filled circles, varying in size as their
judgment of fairness or unfairness. The authors identified the participants gender
and age, family SES, family intactness, and sibship size, their predictor variables, and
fairness and unfairness of each of the techniques in question as the dependent
variables. A second female rater, who was unaware of the studys aims,

independently recorded all ratings for nineteen of the participants, to ensure that the
recorded ratings from the participants were as unbiased as possible (Konstantareas
and Desbois, 2001). Also, the vignettes were presented randomly, and the gender of
the child in each vignette was kept constant for each child participant. However, the
gender of the child in the vignettes was systematically varied within the gender of the
participants. Finally, the children were asked to explain their answers. The
researchers found that female children were more likely than male children to give a
higher score of unfairness to the vignette that described parental differential treatment
of siblings.
Tucker et al., (2003) also found gender oriented reasons for parental
favoritism: gender explained Parental Differential Treatment in parental display of
temporal involvement and affection, and paternal differential treatment was more
common than maternal differential treatment on the basis of sex. Other than gender,
Konstantareas and Desbois (2001) found that family size also might be a reason for
parental favoritism, since children with these demographics rated differential
treatment of siblings unfair more if there was a larger number of siblings in her/his
Perceived (vs Actual! Parental Favoritism
Other researchers have focused on perceived parental favoritism, or the
thought that if a child perceives parental favoritism, s/he may suffer the
consequences, regardless of whether there actually is favoritism in the family or not

(Kowal and Kramer, 1997; and Zervas and Sherman, 1994). According to Kowal and
Kramer, (1997), and their interviews of sixty-one families, parental favoritism may be
childrens creation of the meaning of parental behaviors, and not necessarily the
behaviors themselves, that manipulate childrens reactions. Also, in their quantitative
study of ninety-one college students, Zervas and Sherman, (1994), found that
childrens self observations and evaluation depended upon their perceptions of how
their parents view them.
Mainstream or Pop-Culture Views on Parental Favoritism
In lieu of scholarly attention, a great deal of public attention has focused on
parental favoritism in journalistic accounts. These journalist works tend to depict
parental favoritism as a parents dirty little secret, (Woodman, 2000) or a mothers
greatest taboo. (Maynard, 2003). Most notably, news channels, such as ABC, CBS,
and WebMD Live (on the internet), have gotten involved, bringing experts to their
shows as guests to speak on this ever more concerning matter, from an educated
perspective. This has given the subject a lot of attention, has forced the subject into
the public arena, and asked parents to do some serious introspection.

Theoretical Perspectives in Existing Literature
There have been many different theoretical perspectives that were supported
by or grounded in the research mentioned above. The following theories are only
mentioned once in the existing literature on parental differential treatment:
Attachment Theory claims that child interactions are derived from experience
with their primary caretakers (Brody, Stoneman, and McCoy, 1992);
The Social Information Processing Model of Attribution Theory states that
child perceptions are filtered through past experiences, (Kowal and Kramer,
Equity Theory stressing parental social rewards that each child perceives,
(Brody, Stoneman, and Burke, 1987);
Exchange Theory explains exchanges between parents and children that may
predict parental favoritism, (Brody, Stoneman, and McCoy, 1992);
The Barenbions Developmental Model of Piagetian Theory explains that as
the child gets older they will attainment operational thinking which will give
adolescents the ability to understand parental favoritism, (Alessandri, 1987);
Psychoanalytic Theory stating that children fulfill different needs in different
parents, which is the reason for intra-familial rivalry and thus parental
favoritism, (Barret Singer and Weinstein, 2000 and Brody, Stoneman, and
Burke, 1987);

Psychodynamic Theory explaining that children who are harshly disciplined
may identify with aggressive parents, (Konstantareas, 2001);
Self-esteem Maintenance Theory predicts that sibling perceptions of parental
favoritism will result in conflict, (Brody, Stoneman, and Burke, 1987).
Additionally, The following theories are discussed in multiple studies. Family
Systems Theory claims that parental favoritism will increase with parental conflict
and parent/child alliances, is supported by Barret Singer and Weinsteins (2000)
study. Volling, (1997) and Volling and Elins, (1998) argue that in line with family
systems theory, there are dysfunctional family systems, which are those that create
perverse triangles, or coalitions between parents and children cross-generationally.
This is supported in Vollings 1997 study and Volling and Elins 1998 study. Also,
parental favoritism, or perverse triangles, (as explained above) can cause marital
conflict and sibling rivalry.
In addition, Social Learning Theory is a commonly discussed theory in the
research (Brody, Stoneman, and Burke, 1987; and Brody, Stoneman, and McCoy,
1992; and Zervas and Sherman, 1994). This theory, in reference to parental
favoritism, suggests that there will be lower rates of verbal interaction and prosocial
behavior between siblings, and higher rates of agonistic behavior when there are
higher levels of maternal favoritism, which is supported in a study by Brody,
Stoneman, and Burke, (1987). Brody, Stoneman, and McCoy, (1992) tested this
theory again five years later, and found that children learning particular behaviors in

relationship with their parents, and their parents interactions with others, under social
learning theory, was supported again (Brody, Stoneman, and McCoy 1992).
Interestingly, Zervas and Sherman, (1994) supported Social Learning Theory, finding
that children that reported no favoritism at all in their home had better interpersonal
relations with peers, suggesting that they had learned from their parents to treat peers
Finally, Symbolic Interaction Theory is directly referred to in Zervas and
Sherman, (1994). Symbolic Interaction Theory suggests that childrens feelings about
themselves are derived from childrens perceptions of how their parents view them,
(Gecas, 1982). According to this theory, if a parent openly favors one child over the
other, the non-favored childs self-perception will suffer. Zervas and Shermans
(1994) results were consistent with this theory, finding that self-esteem was
significantly related to parental favoritism.
Socialization Theory
Interestingly, Socialization Theory has not guided research within the
literature on Parental Favoritism. Socialization is the process by which children adapt
to and internalize social norms and values ((Michener, DeLamater, and Schwartz,
1990). When parents have favorites, they teach their children to have favorites, and
then the children grow up to also have favorites. This socialization begins with
favoritism of things such as colors, foods, music, political parties, type of animal, etc.

We are socialized to have favorites regarding everything. At some point this
socialization of favoritism includes favorite household pet, or even favorite parent. In
the previous section, we clearly see that parental favoritism admittedly (by children
and parents alike) exists. Thus, at some point, this socialization of favoritism begins
to include favoring one child over another as well.
Socialization is the way in which individuals learn skills, knowledge, values,
motives, and roles appropriate to their position in a group or society (Michener et al,
1990, pg. 46). For years this has been treated as a uni-directional process, and it
hasnt been until recently that theorists have explored the bi-directional process
concerning childhood socialization. This bi-directional process employs the family,
peers, and formal social institutions, such as school throughout a childs socialization
(Corsaro, 1997). This process can occur through instrumental conditional,
punishment, and observational learning and continues beyond childhood and
adolescence into adulthood.
In reference to favoritism, we can see that the family may socialize the child
(Michener et al., 1990) with the knowledge and values that favoritism is okay, either
by having favorites of their own, or by asking the child to choose a favorite of
something of their own. Peers may influence children and adolescents with the
motive to choose a favorite (for example, the motive to choose a best, or favorite
friend among them) (Corsaro, 1997). And the schools may give their students the
skills to choose favorites, by encouraging decision-making skills (Michener et al.,

1990). At this point, favoritism is now a value or norm within an individual, that can
be built upon. Not only is it okay to choose a favorite color or food, but it is already
okay to choose a favorite friend. At some point an individual may be asked to choose
a favorite parent or a favorite pet, which could translate to choosing a favorite child in
According to Vygotsky (cited in Corsaro, 1997), the individuals
internalization or appropriation of culture, (pg. 15) especially language, is a key
principle of socialization. This is especially interesting to me because the English
language has named favoritism. Thus, we have the ability to conceptualize and
internalize favoritism. A sociolinguistic study of other countries or cultures may give
us insight as to which of them practice favoritism.
Corsaro (1997) renames socialization to prevent the individualistic, forward
looking connotation, and offers the term interpretive reproduction, which is to imply
active innovative and creative contributions to cultural production and change. This
means that not only do families, peers, and schools socialize the child, but the child
interprets that socialization, and reproduces their interpretation of it, thus contributing
to cultural production.
Through socialization, parental values are taken in by the children, and these
children interpret these values, recreating and reproducing them. When a parental
value is favoritism, then the childs value becomes favoritism also. Thus, favoritism
is actually a part of socialization.

The available literature suggests that there are reasons that families practice
parental favoritism, and various consequences for the families that do. The questions
I address are these: What demographic specifications predict or affect parental
favoritism, if any? And, what other variables are valid predictors of parental
Demographic Variables
As the literature suggests, there are many different demographic variables that
can affect parental favoritism. Konstantareas and Desbois (2001) found that girls
were more sensitive to parental differential treatment, and Tucker, McHale, and
Crouter (2003), found that parents have more in common with, and are more
responsible for their same-sex children. Based upon the aforementioned studies, I
hypothesized that male children would be more likely to be favored over female
children by both parents.
Konstantareas and Desbois (2001) found that coming from a large family
predicted an unfair rating for differential treatment of siblings. This may mean that
children from larger families are more sensitive to differential treatment, and thus
affects their perception of parental favoritism. Based upon this, I hypothesized that
there would be more maternal favoritism toward one child over another in families
with a larger number of children.

Tucker, McHale, and Crouter (2003), Volling and Elins (1998), Zervas and
Sherman (1994), as well as McHale and Pawletko (1992) found that birth order of the
children was a predicting variable of parental favoritism. Tucker, McHale, and
Crouter (2003), found that older children had more privileges and chores but less
affection and discipline, while Volling and Elins (1998), found that parents tend to
favor the younger child. McHale and Pawletko (1992) found that parents were more
involved with younger siblings, and Zervas and Sherman (1994), found that children
perceive birth order as one reason for parental favoritism. Based on this literature, I
hypothesized that younger children would receive more maternal favoritism over
older children.
Barrett Singer and Weinstein (2000) found that parents were more likely to
show differential treatment if they were divorced than if they were married. Based
upon this literature, I thought that the number of times a parent was married may also
have an effect on maternal favoritism. I hypothesized that the larger number of
maternal marriages would predict reports of more maternal favoritism.
Barrett Singer and Weinstein (2000) studied Parental Differential Treatment
in European American versus Asian American families, with expectations of finding
reports of more PDT in Asian American families. However, they did not find any
significant differences among the two races. I also think that there will be a
difference among white and non-white families. However, if it is true that children
are socialized into favoritism in the United States, then it is my feeling that white

families will show more maternal favoritism than non-white families, because white
and non-white cultures are so different, even within the United States, that non-white
cultures must influence the way parents deal with their children. Therefore, I want to
test the race effects on maternal favoritism, even though Barrett Singer and
Weinsteins study did not find a significant race effect. I hypothesize that there will
be less favoritism in non-white families.
Much of the literature has been unable to successfully find samples including
demographic variables such as family income and parental educational level, thus the
researchers suggest future research include such variables. Because of this, I included
family income and mothers educational level in my analysis. I hypothesized that
more family income and more maternal education level (only mothers because of
limitations of the data set) will predict less maternal favoritism.
Other Independent Variables
Tucker, McHale, and Crouter (2003) found that a childs gender-type personal
qualities may be able to predict parental favoritism. In their study, parents showed
favoritism toward qualities rated as masculine, such as aggressiveness and
competitiveness as opposed to qualities rated as feminine, such as expressivity and
emotionality, even in girls. Based on this, I hypothesized that the more feminine
qualities a child had, the less maternal favoritism would be shown toward that child.

Crouter, McHale, and Tucker (1999) found that stress can exacerbate the
likeliness of parental favoritism. Thus, I hypothesized that the higher the mothers
stress level the more maternal favoritism she would show.
Kowal and Kramer (1997), and Zervas and Sherman (1994), found that
childrens perceptions of parental favoritism were important. Thus, I hypothesized
that childrens perceptions of maternal warmth would be associated with higher
maternal favoritism.
Barrett Singer and Weinstein (2000) found that when there was less
differential treatment in a family, it predicted more academic achievement. Based on
this finding, I wanted to see if achievement had an affect on maternal favoritism.
Thus I hypothesized that higher verbal achievement would predict lower reports of
maternal favoritism.
Based upon preliminary analysis of the data, I also hypothesized that the more
a mother was involved in her childs social life, the more likely she favored that child.
Finally, I hypothesized that the larger the age difference between the mother and
child, the more likely the mother is to show favoritism. This does not presuppose that
the older the mother the more favoritism, but may imply such. This is only
measuring the age difference between the mother and child. So, for example, if the
mother is forty years old, and there is a twenty year age difference between the
mother and child (mother 40, child 20), versus a thirty-five year old mother with a
thirty year age difference (mother 35, child 5), the thirty-five year old mother is more

likely to show favoritism. Thus, this does not necessarily measure the effect of
mothers age, but the age difference between mother and child.

The National Survey of Children was chosen because it provides a wide range
of data about families, in which not only parents, but also children were interviewed.
Data and Procedure
The National Survey of Children (NSC) includes three waves of data. In the
original sample in 1976, there was a multi-stage stratified probability sample of
households in the continental U.S. containing at least one child. The sampling
produced 2,193 households, from which 2,301 children were interviewed with their
mother (or most knowledgeable parent) in a total of 1,747 households. A follow-up
of the schools attended by the children in the survey was carried out in the spring of
1977. During the time of wave one of the study, children were ages seven to eleven.
Black households were over-sampled to produce, but weights were developed to
adjust for this (Zill, Peterson, Moore, and Furstenbert, Jr., 1992).
The second wave of the NSC focuses on the effects of marital conflict on
children. Children from disrupted families in wave one were sought to re-interview,
totaling 1,423 children completing the second wave. This time there were only 337
Blacks, and the children ranged from age 12 to 17. Telephone interviews were
conducted again with the child and the more knowledgeable parent, by the Institute

for Survey Research at Temple University. There were also questionnaires that were
completed by a teacher, and a subset of interviews that were conducted in person (Zill
et al., 1992).
The final wave, conducted in 1987, included 1,147 interviews completed with
youth. In this sample, since the last wave, some of the children had died and some of
the parents had died, some of the youth were in penitentiaries, and some of the youth
were away (for the military in most cases) for the duration of the study. By this wave
of the study, only 54% of the original respondents selected were interviewed. There
were weights put in place to reduce biases introduced by selective attrition (Zill et al.,
For the purposes of this study, I applied a filter to the sample of families in the
first wave of data to include only those families with two eligible children. In these
families, both children were interviewed, and if there were more than two children,
two focal children were chosen at random, according to the NSC procedure (Zill et
al., 1992).
All but three independent variables are measured in the same way they were
measured by the original researchers from the national data set. The independent
variable for race, (which was originally coded l=black, 2=white, 3=oriental,
4=American Indian, 5=other, and 6=Mexican/Hispanic) was dichotomized as white

and non-white. The independent variable for age difference between mother and
child was constructed by subtracting the childs age from the mothers age. Finally,
the birth order variable was dichotomized to youngest and other.
The dependent variable in the model is maternal favoritism. Favoritism is
measured by adding the following variables (Please see table on pg. 37 for original
Mother allow: watch any TV you want
Mother allow: watch TV whenever want to
Mother allow: play with any friends you want to
Mother allow: snack and eat whatever want
Mother allow: wear any clothes you want
Mother allow: stay up
Mother kiss you or hug you
Mother give you money, extra allowance
Mother buy you something special
A higher value on the D V represents higher level of favoritism.

The following variables were measured as a significant predictor of favoritism
when the significance level is less than .05. There are 7 significant independent
variables in this model. A multiple regression analysis was used to reach these results.
Mothers race is significant at the .044 level. Results showed that white
mothers were less likely to show favoritism than mothers of non-white origin.
Family income is significant at the .013 level. Results indicated that more
family income predicted less maternal favoritism. According to standardized Beta,
for every unit increase in mothers income, there was a 0.080 decrease in favoritism.
Perceived parental warmth is the most significant variable in this model in
predicting maternal favoritism. As expected, at the .000 significance level, the higher
level of perceived parental warmth, the higher level of favoritism. According to
standardized Beta, for every unit increase in perceived parental warmth, there was a
0.510 increase in favoritism.
The politeness of the child is significant at the .020 level. Results showed that
in families in which the child was rated more polite, there was less maternal
favoritism. Thus, the more feminine personal qualities a child has, the less likely s/he
is to be favored. According to standardized Beta, for every unit increase in
politeness, there is a 0.061 decrease in favoritism.

The number of days a mother was nervous, tense, or edgy was significant at a
.009 level. Results indicate that a mother with a higher stress level was less likely to
show favoritism. According to standardized Beta, for every unit increase in mothers
stress level, favoritism decreases .068.
Knowing more of the childs friends is significant at a .016 level. Results
showed that more maternal involvement predicted more maternal favoritism.
According to standardized Beta, for every unit increase in the involvement of the
mother, there was also a 0.067 increase in favoritism.
Age difference between mother and child is significant at the .018 level.
Results indicated that families in which the age difference between the mother and
child was greater, more maternal favoritism. According to standardized Beta, for
every year difference between mother and child, there was a 0.077 increase in
maternal favoritism.
Based on standardized Beta coefficients, one can study the relative importance
of the independent variables, which are as follows from most important to least
important: perceived parental warmth, family income, age difference between mother
and child, mothers stress level, mothers involvement, mothers race, politeness of

Discussing the Results
White mothers in higher family income brackets, tend to have lower scores on
parental favoritism. Race and socioeconomic status have not been widely studied
regarding parental favoritism. This could be because many of the studies have been
qualitative, and the families studied in such research have been white middle class
families. This study differentiated on the basis of white and non-white status with
regard to race. In this respect, it is probable that if non-white races were
differentiated and studied individually, there would be different results for each.
However, Barrett Singer and Weinstein (2000) found that there was no significant
difference in reports of differential treatment by Asian Americans and European
Americans. It was based upon this study that I wanted to test the race variable on
parental favoritism. Race is significant according to the results, but it is significant in
the opposite direction than I predicted. I predicted that non-white mothers would
show less favoritism than white mothers, because of encouraged individualism among
the white culture, and because of communal values among non-white cultures. White
mothers, instead of non-white mothers show less favoritism. White people have

historically been a privileged race in the United States, and recently non-white races
have been able to be more open about anger, suppression, and oppression. Perhaps
non-white races are more open about favoritism also.
Consistent with my hypothesis, however, higher income does predict less
favoritism. It is possible that income has this association because in some cases,
money can cover up feelings of favoritism. For example, if one mother of two
daughters sees a present that she would like to get for one of them, but has enough
money to get two, and give one to each, then she never has to admit thinking of one
over the other. On the other hand, a mother in the same situation without money
would have to forfeit buying the present for either daughter, and thus doesnt have
that to conceal her favoritism. While income has not been studied at all, I speculate
this is because, as mentioned before, there was not enough variation in samples in
previous studies to draw from. Thus, this result should add to the body of literature
regarding parental favoritism, for other researchers to expand on.
Also consistent with the stated hypothesis higher levels of perceived parental
warmth showed higher levels of favoritism. This also supports the literature
regarding childrens perceptions of parental behaviors versus the behaviors
themselves, as an indicator of favoritism.
In addition, polite children were less likely to be favored, and stressed mothers
were less likely to show favoritism. Again, the literature is supported here. Personal
qualities of the mother and the child are predictors of favoritism. It is possible that a

polite child has already adapted to being more polite as an adjustment to being the
non-favored. By this I mean that a polite child may be attempting to win over the
favor of the parent, or gain any approval from adults for that matter. In addition, it
could also indicate a child that acts out gets more attention, and perceives that as
favoritism. The mothers stress level, on the other hand, may be an indication that
when a parent is edgy or stressed, it is directed at everyone equally, thus any
preference is less visible. In addition, stress indicates that the parent is stressed about
something, and preoccupied with thought of that something, and thus not consciously
giving any thought to a favorite child.
Also consistent with our hypotheses, results indicated that a larger age
difference between the mother and the child results in more parental favoritism. One
could speculate that this could be because a child that is bom later in a mothers life is
less likely to be a mistake, or more wanted, or even medically sought after.
Therefore, the parent is not only not resentful of the child, but appreciates it, because
it was a chosen situation, instead of a forced one. One might also speculate that a
larger age difference between the mother and the child may indicate a parent with
more experience, and more knowledgeable with regard to parenting. A parent with
more experience in child rearing may feel more comfortable with themselves, and
thus more comfortable with the child at that point in their life, and thus favor that
specific child.

In addition, and consistent in reference to our hypotheses, results show that
mothers who favor a child tend to know the childs friends. This was not surprising
because I expected that a mother who favors a child would also want to be involved
in that childs social network, and perhaps even be accepted by that childs friends. It
also implies that the parent would show more interest in a favored childs friends,
because the parent would be more involved with the favored childs social network,
and less involved with the non-favored childs social network.
Surprisingly, in this analysis, the childs gender, the mothers education level,
family size, birth order, and the number of times the mother had been married were
not significant. There are many possible explanations for this lack of significance.
One explanation might be that since we only measured maternal favoritism, and some
studies show that parents have more in common with same-sex children, (Tucker,
McHale, and Crouter, 2003), perhaps mothers are more balanced between males and
females, and thus the gender of the child was not significant in this study. This may
have also been a factor regarding the birth order variable. Mothers may tend to be
more balanced with regard to favoring first-born vs. youngest, but if we had measured
paternal favoritism, this may have been a significant variable. I speculate that the
number of marriages may not have been significant because if a pattern of favoritism
already exists before a mother remarries, then the new marriage may not change that
pattern. For example, if a mother has two children from her first marriage, and
already favors one child over the other, then it is not likely that any additional

marriages are going to change that already existing pattern. However, introducing a
third child from one of those new marriages as a new variable to this existing pattern,
creates yet another dynamic that could affect the existing pattern. At that point, it is
still the third child that affected the pattern, not the new marriage.
The limitations of this study are obvious. This study was limited by the
procedure and questions used by the original researchers. Dealing with secondary
data creates problems of operationalization, because I was limited to the questions
that the previous researchers chose to ask, which also causes problems of validity. In
other words, because this is a secondary analysis, I was unable to choose the
questions asked in this survey. In addition, fathers were under-represented in this
sample, which was also an uncontrollable limitation of this data set. Because of this
under-representation, I was only able to use measure maternal favoritism. Finally, all
of this measurement is based on childrens perceptions of maternal favoritism, and
therefore when I refer to favoritism, it may be understood as perceived maternal
favoritism. However, the need to learn more about this topic exceeds the limitations
of this data set. These data have allowed me to analyze a large sample, and add to the
knowledge on this topic, as very few researchers have attempted to analyze a national
data set to learn more about parental favoritism.

Suggestions for Future Research
Due to the limitations of a study of this nature, future research should continue
to concentrate on comparing and contrasting families from different cultural contexts,
within and outside of the United States. There have been many more qualitative
studies on this subject than quantitative. A questionnaire designed specifically
regarding parental favoritism could benefit this area of study (instead of relying on a
questionnaire developed to measure sibling rivalry, as many previous studies have
done). Ideally, this questionnaire would measure derivations of, reasons for, and
consequences of parental favoritism, not only for the child, but also for the parents.
Another question to address might be whether parents who come from
families that practiced parental favoritism are more likely to practice it as well, or if
they are more likely to rebel from it. Finally, all researchers to this point have
concentrated on the effects of parental favoritism on children, and I believe that the
parents that practice favoritism are possibly affected by their own actions as well. I
suggest that future research on this topic look into how parental favoritism affects the
parents in addition to how it affects the children.

Table Al: The Wordings, Codings, Means, and Standard Deviations of
Parental Favoritism Variables
Mean Deviation
Dependant Variable: Favoritism (0=least favored,4.14 1.732
and 9=most favored)
manytv mtvanytm mfriends msnack mclothes mstayp mkiss mmoney mpresent Mother Allow: Watch Any TV You Want Mother Allow:Watch TV Whenever Want To Mother Allow: Play With Any Friends You Want Mother Allow: Snack And Eat Whatever Want Mother Allow: Wear Any Clothes You Want Mother Allow: Stay Up Mother Kiss You Or Hug You Mother Give You Money, Extra Allowance Mother Buy You Something Special (l=yes and 0=no)
Gender childsex Sex of child
(l=boy, 2=girl)
Mothers Education Level
edumo Mothers Education 2.86 1.050
Race white (l=primary only, 2=some secondary, 3=HSgrad, 4=some college, 5=college grad) White (l=white, 0=non-white)
Family Size numchld Total # of children 3.68 2.037 (1-15)

Mean Std.
Family Income Deviation
incomexl Income (1<$5000, 2=$5000-<$ 10000, 3=$10000- <$15000,4=$ 15000-<$25000,5=$25000 & >) Childs Age 3.04 1.242
childagec Age of child (6-12) Birth Order bo_young Birth Order (l=youngest, 0=all other) Achievement 8.97 1.583
academic How good student is child (l=best,..and 5=near bottom) 2.26 .984
cverbal Age/Series Adjusted Vocabulary Score (5-86) Maternal Involvement 50.11 10.122
cnfknown # childs friends you know (l=all, 2=most, 3=about half, 4=few, 5=none) Childs Personal Qualities 1.66 1.068
cpolite Child is polite, helpful, considerate (l=not at all, 2=little, 3=somewhat, 4=pretty much, 5=very much, 6=exactly like) Maternal Stress Level 4.50 1.181
tensemo Days mother nervous/tense/edgy (l=very often, 2=fairly often, 3=occasionally, 4=hardly ever) Perceived Maternal Warmth 2.63 .924
warmthm Perceived parental warmth (1 l=least warm,..and 69=warmest) Number of Maternal Marriages 50.04 9.698
marnumma Number of marriages of mother (0=0 previous marriages, 2=two, 3=three, 4=four) Age Difference between Mother and Child 1.21 .538
agedijf Age of mother age of child (13-71) 26.2482 6.89211

Table A2: Unstandardized and Standardized Coefficients of Dependent
and Independent Variables of Favoritism
Dependent Variable: Favoritism constructed by adding the following: (1) Mother
allow: watch any TV you want (2) Mother allow: watch TV whenever want to (3)
Mother allow: play with any friends you want to (4) Mother allow: snack and eat
whatever want (5) Mother allow: wear any clothes you want (6) Mother allow: stay
up (7) Mother kiss you or hug you (8) Mother give you money, extra allowance (9)
Mother buy you something special
Independent Variables
Sex of child
Mothers Education
Total # of children
Age of child
Birth Order
How Good Student Is
# Childs Friends You Know
Child Is Polite, Helpful, Considerate
Age Series Adjusted Vocab Score
Days mother nervous/tense/edgy
Perceived parental warmth
Number of marriages of mother
Age of mother age of child
Unstandardized Standardized
Coefficients Coefficients
-.118 -.034
-.041 -.025
-.244* -.062*
.000 .000
-.112* -.080*
.027 .024
.130 .036
-.082 -.048
.110* .067*
-.092* -.061*
-.009 -.061
-.130** -.068**
.091** .510**
-.097 -.025
.020* .077*
R square = .32
Notes: : Alpha = .05
Alphas .01

Alessandri, S., Wozniak, R The Childs Awareness of Parental Beliefs concerning
the Child: A Developmental Study. Child Development. 1987 v. 58
p. 316-323
Barret-Singer, A., Weinstein, R. Differential Parental Treatment Predicts
Achievement and Self-Perceptions in Two Cultural Contexts. Journal of
Family Psychology. 2000 v. 14 n. 3 p. 491-509
Brody, G., Stoneman, Z., Burke, M. Child Temperaments, Maternal Differential
Behavior, and Sibling Relationships. Developmental Psychology. 1987 v. 73
n. 3 p. 354-362
Brody, G., Stoneman, Z., McCoy, K. Associations of Maternal and Paternal Direct
and Differential Behavior with Sibling Relationships: Contemporaneous and
Longitudinal Analyses. Child Development. 1992 v. 63 p. 82-92
Brody, G., Stoneman, Z., McCoy, K. Parental Differential Treatment of Siblings and
Sibling Differences in Negative Emotionality. Journal of Marriage and the
Family. August 1992 v. 54 p. 643-651
Corsaro, William. The Sociology of Childhood. Thousand Oaks: Pine Forge Press,
Crouter, A., McHale, S., Tucker, C. Does Stress Exacerbate Parental Differential
Treatment of Siblings? A Pattern-analytic Approach. Journal of Family
Psychology. June 1999 v. 13 n. 2 p. 286-299
Gecas, V. The Self-Concept. Annual Review of Sociology. 1982 v. 8 p. 1-33
Harris, K., Morgan, S. Fathers, Sons, and Daughters: Differential Paternal
Involvement in Parenting. Journal of Marriage and the Family. August 1991
v. 53 p. 531-544
Konstantareas, M., Desbois, N. Preschoolers Perceptions of the Unfairness of
Maternal Disciplinary Practices. Child Abuse & Neglect. 2001 v. 25 p. 473-

Kowal, A., Kramer, L. Childrens Understanding of Parental Differential
Treatment. Child Development. Feb. 1997 v. 68 n. 1 p. 113-126
Lang, S. Mothers Have Favorite Child. Human Ecology. Winter 1998, v. 26 i. 1
p. 23
Maynard, J. The Favorite Child. Good Housekeeping. Mar. 2003 v. 236 i. 3 p. 101-
McHale, S., Pawletko, T. Differential Treatment of Siblings in Two Family
Contexts. Child Development. 1992 v. 63 p. 68-81
McHugh, C. Do You Play Favorites: A family secret thats more common than you
think? Family Life. Sept/Oct 1996
Michener, H., DeLamater, J., Schwartz, S. Social Psychology: Second Edition. San
Diego: Harcourt Brace Jovanovich, Publishers, 1990
Strong, B., DeVault, C., The Marriage and Family Experience: Sixth Edition.
Minneapolis/St. Paul: West Publishing Company, 1995
Tucker, C., McHale, S., Crouter, A. Dimensions of Mothers and Fathers
Differential Treatment of Siblings: Links with Adolescents Sex-Typed
Personal Qualities. Family Relations. Jan. 2003 v. 52 n. 1 p. 82-89
Volling, B. The Family Correlates of Maternal and Paternal Perceptions of
Differential Treatment in Early Childhood. Family Relations. 1997 v. 46 n. 3
p. 227-236
Volling, B., Elins, J. Family Relationships and Childrens Emotional Adjustment as
Correlates of Maternal and Paternal Differential Treatment: A Replication
with Toddler and Preschool Siblings. Child Development. Dec. 1998 v.69
n. 6 p. 1640-1656
Woodman, S. When Parents Play Favorites. Family Life. October 2000 p. 62-66
Zervas, L., Sherman, M. The Relationship Between Perceived Parental Favoritism
and Self-Esteem. Journal of Genetic Psychology. 1994 v. 155 n. 1 p. 25-33

Zill, N., Peterson, J.L., Moore, K.A. and Furstenberg, F.F. Jr. (1992) 1976-1987
National Survey of Children: Waves 1, 2, and 3 (Child Trends, Inc.) (Data Set
6-12, Crutchfield, J.H., Schwartz, S.M., & Card, J J., Archivists) [machine-
readable data file and documentation]. Washington, D.C.: Child Trends. Inc.
(Producer). Los Altos, CA: Sociometrics Corporation, American Family Data
Archive (Producer & Distributor).