Family Patterns of Perfectionism: An Examination of College Students ...

This article was downloaded by: [Duke University Libraries] On: 30 May 2012, At: 09:09 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Personality Assessment Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hjpa20

Family Patterns of Perfectionism: An Examination of College Students and Their Parents Angela Z. Vieth & Timothy J. Trull Available online: 10 Jun 2010

To cite this article: Angela Z. Vieth & Timothy J. Trull (1999): Family Patterns of Perfectionism: An Examination of College Students and Their Parents, Journal of Personality Assessment, 72:1, 49-67 To link to this article: http://dx.doi.org/10.1207/s15327752jpa7201_3

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/termsand-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages

Downloaded by [Duke University Libraries] at 09:09 30 May 2012

whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.


JOURNAL OF PERSONALITY ASSESSMENT, 72(1), 49-67 Copyright O 1999, Lawrence Erlbaum Associates, Inc.

Family Patterns of Perfectionism: An Examination of College Students and Their Parents Angela Z. Vieth and Timothy J. Trull Department of Psychology University of Missouri-Columbia

We examined patterns of perfectionism among college students and their biological parents in a sampleof 188undergraduates from intact families.Ratings (self vs. other) showed the greatest degree of convergence when daughters were either the target or the rater. Levels of self-oriented perfectionism in students were positively associated with the levels characterizingthe same-sex parent, but unrelated (father-daughter) or negatively related (mother-son) to the levels characterizing the opposite-sex parent. Finally, parents' other-oriented perfectionism was not significantly related to students' socially prescribed perfectionism. The construct of perfectionism has received an increasing amount of attention in the psychological literature over the last 40 years. Although the basic definition of perfectionism-the tendency to set extremely high standards-has changed little over this time period, the overall desirability, or healthiness, of perfectionistic strivings has remained a topic of debate. Further, many attempts have been made to distinguish harmful aspects of this construct from those that may be beneficial or benign (Hamachek, 1978; Missildine, 1963; Slade, Newton, Butler, & Murphy, 1991). In recent years, perfectionism has been linked with a number of physical and psychological problems, both mild and severe: sleep difficulties (Lundh, Broman, Hetta, & Saboonchi, 1994); irritable bowel syndrome, ulcerative colitis, Munchausen syndrome, alcoholism, and writer's block (Pacht, 1984); procrastination (e.g., Sorotzkin, 1985); Type A behavior (Flett, Hewitt, Blankstein, & Dynin, 1994); eating disorders (Ordman & Kirschenbaum, 1986; Slade, 1982) and unhealthy eating attitudes (Waller, Wood, Miller, & Slade, 1992); suicidal ideation-potential (Hewitt, Flett, & Turnbull-Donovan, 1992; Hewitt, Flett, & Weber, 1994); irrational beliefs (Flett, Hewitt, Blankstein, & Koledin, 1991) and a


50

VIETH AND TRULL

range of specific fears (Blankstein,Flett, Hewitt, & Eng, 1993);erectile dysfunction (Quadland, 1980); troubled personal relationships (Burns, 1980); and personality disorders (Broday, 1988). The condition most strongly associated with perfectionism, and the one that has received the majority of the attention in the literature, is depression (e.g., Blatt, 1995). Investigators have long surmised that depressed individuals are prone to set "stringent criteria for self-evaluation" (Rehm, 1977, p. 793), a tendency that is marked in perfectionists. However, research does not support the view that high standards are necessarily linked to depression (Carver & Ganellen, 1983; Kanfer & Zeiss, 1983; Pirot, 1986). Rather, perfectionism may be more accurately regarded as a predisposing factor for depression (Burns, 1980; Flett, Hewitt, & Dyck, 1989; Hewitt & Dyck, 1986; Hewitt & Flett, 1993), or as "one element in a diathesis-stress model of depression" (Hett et al., 1989, p. 732).

ETIOLOGICAL THEORIES OF PERFECTIONISM Given the significance of perfectionism, several writers have theorized about the origins of the construct. In each case, perfectionism is viewed as an interpersonal style adopted primarily in response to interactions with one's caregivers. Missildine (1 963), for instance, believed that perfectionism is rooted in "persistent parental demand" (p. 94) and the withholding of parental acceptance. Hollender (1965) presented a very similar formulation, in which parents promote perfectionism in their children by being exacting and by responding to their children in aconsistently conditional and dichotomous manner. In both formulations, "budding" perfectionists come to believe that if they work hard enough, their parents will eventually grant them full acceptance. Hamachek (1978) distinguished between "normal" and "neurotic" perfectionism. According to this theorist, normal perfectionism is the result ofpositive modeling (as when one emulates parental strivings) or negative modeling (as when one reacts against parental disorganization or underachievement).Neurotic perfectionism, on the other hand, arises either from environments of conditional approval (similar to the formulations of Hollender, 1965,and Missildine, 1963)or conditions of nonapproval or inconsistent approval. In any case, the child adopts personal standards so high that their attainment should please anyone. Barrow and Moore (1983) postulated that perfectionistic thinking is always the result of conditional positive approval, or "the fusion of self-worth with achievement" (p. 613). Furthermore, these authors outlined several conditions that might lead to this type of fusion: overt and harsh parental criticism; more subtle criticism, which parents might communicate by stating their standards and expectations; the lack of clearly communicated standards (similar to Hamachek's, 1978,condition of nonapproval); or the observation of perfectionistic behavior in parents (Barrow & Moore, 1983). Finally, Burns


FAMILY PATTERNS OF PERFECTIONISM

51

(1980) asserted that the origins of perfectionism lie, at least in part, in interactions with perfectionistic parents. Within such a relationship, the child is positively reinforced for excellent performance. However, when he or she fails at an endeavor, his or her parents view the failure as a reflection of their childrearing skills. Rather than reassuring and comforting the child, they react with disappointment, anxiety, or agitation. As aresult, the child begins to dread all mistakes, and his or her own reactions further reinforce the behavior of the parents. In this brief review on the etiology of perfectionistic tendencies, an impressive degree of convergence may be observed across theorists: The child adopts perfectionistic tendencies either by modeling himself or herself after parents or by reacting to parental nonapproval, inconsistent approval, or conditional approval. Unfortunately, none of the theorists cited has speculated about the relative influence of each parent in the solidification of this trait. Only three empirical studies to date have been published on the development of perfectionism (Flett, Hewitt, & Singer, 1995; Frost, Lahart, & Rosenblate, 1991; Parker, cited in Blatt, 1995). Flett et al. (1995) focused on the relation between parental authority styles (i.e., authoritarian, permissive, and authoritative) and students' perfectionism scores. Of primary relevance to this study, Flett et al. found that male students' socially prescribed perfectionism scores (reflecting the tendency to see others as setting unrealistically high standards) were significantly related to both mothers' and fathers' levels of authoritarianism (restrictive, punitive, and overcontrolling). The major limitation of this study was that all ratings (including ratings of parents) were made by the students. Frost et al. (1991) hypothesized that perfectionism among parents is associated with perfectionistic tendencies in their children. To assess perfectionism, Frost et al. used the Frost Multidimensional Perfectionism Scale (FMPS; Frost, Marten, Lahart, & Rosenblate, 1990). This instrument yields an overall perfectionism score as well as scores on six subscales: Concern Over Mistakes, Personal Standards, Parental Expectations, Parental Criticism, Doubts About Actions, and Organization (Frost et al., 1990). The FMPS was administered to undergraduate women living in a small dormitory and, via mail, to the biological parents with whom the students indicated they had lived for a "significant amount of time" while growing up (Frost et a]., 1991).Forty-one of the undergraduate women completed the FMPS and gave the researchers permission to contact their parents. A total of 31 of the 41 mothers contacted completed and returned the FMPS, as did 25 of the 38 fathers listed. Correlations were computed between the FMPS scores of the undergraduate women and each of their parents. These analyses revealed a significant correlation between the overall perfectionism scores of mothers and daughters, suggesting that women with perfectionistic mothers are more perfectionistic themselves. However, no significant relation was obtained between the overall perfectionism scores of fathers and daughters. This study was significant in that it empirically supported what had only been speculated previously: Parental


52

VIETH AND TRULL

perfectionism (specifically, perfectionism in mothers) is related to perfectionism in children (specifically, daughters). Unfortunately, because the sample used was composed entirely women, we may not generalize this finding to offspring of both sexes. Nor can we tease apart the relative influences of the primary caregiver and the same-sex caregiver. In an unpublished study, Parker (cited in Blatt, 1995) investigated the relation between children's and parents' perfectionism scores in a large sample of gifted sixth grade children. Parker found that mothers' and fathers' total scores on the FMPS explained 21% and 11%, respectively, of the variance in their children's scores on the same measure. Interestingly, mothers' scores accounted for a greater proportion of the variance in daughters' scores than in sons' scores (29% as opposed to 20%), whereas the reverse pattern was observed for fathers (fathers' scores accounted for 15% of the variance in sons' scores but only 5% of the variance in daughters' scores). These three studies suggest that perfectionistic tendencies in parents may be related to those reported by their offspring. This study had three main goals. First, we assessed the relations between students' self-reports of perfectionism scores and parents' ratings of students on these dimensions, as well as the relations between parents' self-ratings and students' ratings of their parents. In this way, we estimated the amount of convergence between students and parents when focusing on the same targets.' An additional goal of this study was to examine the association between parents' and students' levels of perfectionism. We were interested in whether (and how) parents' tendency to set unrealistically high standards for themselves was related to this same tendency in their offspring. In the Western world, it is still likely to be the case that individuals in their formative years have much greater exposure to their mothers than to their fathers (Belsky, Rovine, & Fish, 1989; Parke, 1995; Whiting & Edwards, 1988). If the perfectionism scores obtained for young adults correspond more closely with those of their mothers, as opposed to their fathers, support would be granted to the notion that individuals' levels of perfectionism are more closely associated with the levels characterizing their primary caregiver^.^ Such a finding might be explained by differences in the quantity and/or quality of one's exposure to each parent while growing up. However, if a different pattern of correlations results for male and female participants such that scores obtained by individuals of a given sex correspond more strongly to those obtained by their l ~ l t h o u ~here h we focus only on the relations between self-other perceptions, other relations could be explored as well (Kenny, 1994). 21n this study, we assumed that the biological mother from intact families served as the primary caregiver. Supporting this assumption, observational studies in the United States and elsewhere document that mothers are much more involved in the care of children than fathers, regardless of the age of the child (i.e.,from infancy to adolescence; Belsky, Rovine, &Fish, 1989; Parke, 1995; Whiting & Edwards, 1988).


53

same-sex parents, support would be granted to the notion that individuals' levels of perfectionism are more closely associated with the levels characterizing parents of the same sex. This type of finding might be explained by the phenomenon of


same-sex modeling.

Finally, we examined the relation between parents' tendency to set exacting standards for others and students' perceptions that others place unreasonable expectations on them. If such a relation exists, support would be granted to the proposal that parents' high expectations of others may, in turn, influence students' reports that others' expectations are too high.

METHOD Participants Participants were 303 studentsrecruited from an introductory psychology course at a large, midwestern university. Participants attended one of several 1-hr,group laboratory sessions offered in exchange for a required experimental credit. Participants were asked to confirm that they were from an intact family (i.e., reared by both biological parents together until the participant moved out or left for school). Intact families were used to ensure that all parents' reports were based on 18 years of exposure to the student. Of the original 303 participants, 4 participants were dropped after it was determinedthat they were not from intact families, and 3 others were dropped after it was learned that they were siblings of other study participants. Thus, the reduced sample included 296 participants. Also serving as participants in the study were biological mothers and biological fathers of the students, whose participation was solicited by mail. The parents of seven students were not contacted because either the student withheld permission ( n = I), or the parents resided outside of the United States (n = 6). In addition, one student did not know the current whereabouts of his father, and one other indicated that his father was deceased. Altogether, attempts were made to contact the mothers of 289 participants and the fathers of 287; of these, 212 mothers (73.4%) and 194 fathers (67.6%) consented to participate, resulting in an overall response rate of 70.5%. In the case of 188 students, both parents provided data for the study. Unless otherwise indicated, the analyses presented in this article utilize data from these 188 families only. See Table 1 for the demographic characteristics of students in this restricted sample. Of the 188 students in this subset, 128 were women and 60 were men. The majority (66.3%) were freshmen, and their ages ranged from 17 to 38 years (M = 18.95 years, SD = 1.97).The sample was predominantly White (94.7%) and single (98.9%) The only notable difference in the demographic information provided by the students in the total and restricted samples was a greater propor-

54

VIETH AND TRULL

TABLE 1 Demographic Characteristics of the Sample


Variable

Women

Men

Total Sample

Age n M SD Race n White African American Asian Hispanic Other Religion n Catholic Protestant Jewish Other None Marital Status n Single Married Year in College n Freshman (9%) Sophomore (%) Junior (%) Senior (%) Graduate (%) Note. ns may differ slightly for each variable, because some participants failed to provide this information.

tion of White participants, and smaller proportions of Asian and African American students in the sample with both parents responding. Of the 188 sets of parents in the restricted sample, demographic data were provided by 173 mothers and 169 fathers. The mothers ranged in age from 36 to 70 years (M = 46.29 years, SD = 5.13) and had completed an average of 14.50years of education (SD= 2.26); the fathers ranged in age from 37 to 69 years (M = 48.38 years, SD = 5.62), with an average education of 15.47 years (SD= 2.58). The vast majority (98.4%) of parents were still married, with an average length of marriage of approximately 25 years ( M = 25.07, SD = 4.37). The average number of children was 2.8 1 (SD= 1.13) including the student participant.

FAMILY PA'ITERNS OF PERFECTIONISM

55


Measures Multidimensional Perfectionism Scale (MPS). The MPS (Hewitt & Flett, 1989) is a 45-item self-report inventory composed of three subscales of 15 items each. Each subscale corresponds to one of three perfectionism dimensions conceptualized by Hewitt and Flett (1989). The Self-oriented Perfectionism subscale measures the tendency to set unrealistically high standards for oneself and judge oneself on the basis of these harsh dictates; items include "It makes me uneasy to see an error in my work" and (reverse-coded)"I do not have to be the best at whatever I am doing." The Other-Oriented Perfectionism subscale reflects the extent to which one sets exacting standards for others and subjects them to stringent evaluation; items include "The people who matter to me should never let me down" and (reverse-coded) "I do not expect a lot from my friends." The Socially Prescribed Perfectionism subscale measures the extent to which one feels that others place unreasonable standards on one's behavior and evaluate one harshly; items include "My family expects me to be perfect" and (reverse-coded)"Those around me readily accept that I can make mistakes too." Participants rate their agreement with the items of the MPS on a 7-point Likert scale; a higher score reflects greater levels of a particular dimension of perfectionism. Several of the items on the MPS are reverse-coded. In general, the MPS has good psychometric properties and appears to be both reliable and valid (Hewitt & Flett, 1991; Hewitt, Flett, Turnbull-Donovan & Mikail, 1991). Hewitt, Flett, and their colleagues have employed the MPS in numerous studies. Of relevance to this study, significant correlations have been reported between self- and observer-ratings in both clinical and student populations, suggesting that "perfectionism is salient in interpersonal contexts" (Hewitt & Flett, 1991, p. 460).

Procedure At the beginning of the laboratory session, students were asked to provide the names and addresses of their biological parents such that they might be solicited for participation in the second stage of the study. In addition, assurances were made regarding the confidentiality of all participants' responses. Each student completed a number of pencil-and-paper measures, including the MPS. In addition, students were asked to rate each of their parents using modified forms of the MPS. For these modified versions of the MPS, the wording of the items was changed to the third person (i.e., I was changed to she or he, etc.) and an additional instruction (e.g., "Please use this form to rate your mother") was added to the top of the sheet. As indicated earlier, parents were contacted through the United States Postal Service. Each set of parents was mailed a packet including acover letter; informed con-


56

VlETH AND TRULL

sent forms; demographic data sheets; original (for rating self) and modified (for rating the student) forms of the MPS; and pre-addressed, stamped envelopes in which to return the consent forms, the data sheets, and the completed inventories. Mothers and fathers were provided with separate return envelopes to help ensure a confidential response. As an inducement to participate, parents were informed that upon completing and returning the questionnaires, they would be entered into a drawingfor a $50cash prize to be awardedat theend of the study.Eachparent was instructed to fill out all materials independently (i.e., without consulting either his or her spouse or the student) and to return these materials within 3 weeks. After 3 weeks, reminder postcards were sent to each parent who had not yet responded. Because the goal was to obtain as many complete sets of parent ratings as possible, a second batch of reminder postcards was sent after 3 additional weeks to only those individuals whose spouses had already responded. To examine whether the perfectionism ratings of students with both parents responding (the restricted sample) differed from the perfectionism ratings of students with zero or one parent participating, a series o f t tests was conducted. No group differences were found for any of the perfectionism scores examined (all ps > .05).

RESULTS Psychometric Properties of the MPS

Means, standard deviations, and comparisons with norms. The means and standard deviations for each self-rated MPS subscale are presented in Table 2 for male and female students and their parents. In general, these values are very TABLE 2 Students' and Parents' Scores on the Subscales of the Multidimensional Perfectionism Scale

Parents

Students Women" Measure Self-oriented Other-Oriented Social Prescribed

Menh

Mothersc

Fathers"

M

SD

M

SD

M

SD

M

SD

71.14 59.32 51.16

15.11 11.64 15.01

71.25 62.29 57.54

14.47 10.15 12.35

69.09 58.93 50.56

17.19 11.90 15.25

69.82 62.31 51.40

15.11 11.46 13.24

Note. All values were derived from self-ratings. " n = 128. ' n = 6 0 . ' n = 184.*n= 181.


57


similar to the community norms presented in Hewitt et a]. (1991) as well as the means (student samples) reported in Hewitt and Flett (1991, Studies 1-4).

Internal consistency. Internal consistency reliabilities (Cronbach's a) for all three subscales of the MPS were computed. The reliability coefficients for the MPS subscales were high for both parents and students, ranging from .88 to .91 for self-oriented perfectionism, .76 to .81 for other-oriented perfectionism, and .86 to .88 for socially prescribed perfectionism. Intercorrelations among the subscales of the MPS. Students' scores on the subscales of the MPS were positively and significantly intercorrelated. Self-rated, self-oriented perfectionism correlated .48 with other-oriented perfectionism and .44 with socially prescribed perfectionism, and other-orientedperfectionism correlated .37 with socially prescribed perfectionism (all ps < .001). Relations of similar magnitude were observed among both mothers' and fathers' selfreported perfectionism scores (all ps < .001). The present intercorrelations are somewhat larger than those reported by Hewitt and Flett (rs ranging from .25 to .40) in another sample of college students (Hewitt & Flett, 1991; Study 1). However, Hewitt and Flett (1991) have explained that, due to their common "focus on the attainment of standards" (Hewitt & Flett, 1991, p. 458), the perfectionism dimensions should be expected to overlap.

Convergence of Self- and Observer Ratings of Self-oriented Perfectionism Table 3 presents results indicating the relative degree of agreement between students and parents regarding the respective targets' standing on the three dimensions of perfectionism. The top panel presents correlations between students' self-ratings and ratings of the student by each parent. There appeared to be greater convergence between parents' ratings of daughters' perfectionism and daughters' self-reports than observed for sons. However, it is interesting to note that for sons, the highest convergent coefficients were obtained for mothers' and fathers' ratings of sons' socially prescribed perfectionism. The bottom half of the table presents correlations between parents' self-reported perfectionism scores and students' ratings of the parents. In general, daughters' ratings corresponded more highly with parents' self-reports than did sons' ratings. The one notable exception was that sons' ratings of fathers' self-oriented perfectionism correlated more highly with fathers' self-reports than did daughters' ratings.

58

VlETH AND TRULL

TABLE 3

Correlations Between Self-Ratedand Observer-Rated Multidimensional Perfectionism Scale Scores


Self-oriented Perfectionism

Other-Oriented Perjectionism

Socially Prescribed Perfectionism

Daughters' self-rated scoresa Mothers' ratings of daughters Fathers' ratings of daughters Sons' self-rated scoresh Mothers' ratings of sons Fathers' ratings of sons Daughters' ratings of parentsd Mothers' self-rated scores Fathers' self-rated scores Sons' ratings of parentsh Mothers' self-rated scores Fathers' self-rated scores Note. These analyses are restricted to data from families in which all three members (student, mother, and father) supplied self-ratings of self-oriented perfectionism. * p < .05.**p < .01. ***p < .W1.

Association Between Parents' and Students' Levels of Perfectionism To assess the relation between parents' and students' self-reportedlevels of perfectionism, hierarchical regression analyses were conducted. First, we evaluated whether parents' self-oriented perfectionism scores were related to students' self-oriented perfectionism scores and whether this was moderated by the biological sex of the student. In this model, sex of student was entered first, followed by mothers' and fathers' self-oriented perfectionism scores (entered simultaneously), and finally the two interaction terms (entered simultaneously).Prior to computing cross-product terms and entering variables into the model, all predictor variables were centered (in this case, standardized) to remove nonessential collinearity (Aiken & West, 1991). The results presented in Table 4 indicate that the block of mothers' and fathers' self-oriented perfectionismscores was a significant predictor (Step 2), and these relations were qualified by a moderator effect involving student biological sex (Step 3). Specifically, the relations between both mothers' and fathers' self-oriented perfectionism scores respectively and students' self-oriented perfectionism scores varied as a function of sex of the student. Table 5 presents correlations that indicate the nature of these interaction effects. It was proposed that either the correlations between the self-oriented perfectionism scores of students and mothers would be significantly greater than the cor-


59

TABLE 4 Hierarchical Multiple-RegressionResults: Predicting Students' Self-oriented Perfectionism From Parent's Self-oriented Perfectionism Scores


Order (J'Entry qf' Predictor Variables

Model df

RZTotal

A R2

AF

F Overall

1, 175 3, 173 5. 171

.00

.00 .04 .11

0.05 3.39* 11.38**

0.05 2.27 6.08**

Step 1: Students' Sex Step 2: Mothers' SOP and father's SOP Step 3: Interactions Students' Sex x Mothers' SOP Students' Sex x Fathers' SOP

.04 .15

Note. SOP = self-oriented perfectionism. *p < .05. **p < ,001.

TABLE 5 Correlations Between Students' and Parents' Self-Rated Multidimensional Perfectionism Scale Scores Self-oriented Perfectionism

Daughters' scores" Mothers Fathers Son's scoresh Mothers Fathers p

p

p

p

p

p

p

Other-Oriented Perfectionism

Socially Prescribed Perfectionism

.31** .00 -.28* .46** p

p

p

p

p

p

~

-

-

-

-

-

-

-

-

-

-

-

-

--

Note. These analyses are restricted to data from families in which all three members (student, mother, and father) supplied self-ratings of self-oriented perfectionism. "n= 119. 'n = 58. *p < .05. **p < ,001.

relations between the scores of students and fathers, regardless of student sex (the primary caregiver hypothesis), or the correlations between the self-oriented perfectionism scores of students and their same-sex parents would be greater than the correlations between the self-oriented perfectionism scores of students and their opposite-sex parents (the same-sex modeling hypothesis).3 As can be seen, the data provided support for the latter hypothesis. The correlation between daughters' and mothers' self-oriented perfectionism scores (r = .31) was significantly larger than the correlation between daughters' and fathers' scores, r = .00; t ( l 1 6 ) = 2.54, p < .05. Moreover, the correlation between sons' and mothers' self-oriented perfectionism scores (r = -.28) was significantly smaller than the correlation between 3 ~ o these r and other comparisons of dependent correlations, a t statistic with n - 3 degrees of freedom was calculated (Cohen & Cohen, 1983).


60

VlETH AND TRULL

sons' and fathers' scores, r = .46; t(55) = -4.48, p < .01. Thus, we may conclude that students' levels of self-oriented perfectionism were associated with the levels characteristic of their same-sex parents. Further examinationof Table 5 suggeststhat the pattern observed with respect to self-oriented perfectionism was not demonstrated with regard to either other-oriented perfectionism or socially prescribed perfectionism. Although mothers' and daughters' scores on the socially prescribed perfectionism subscale were significantly related, mothers' and daughters' other-orientedperfectionism scores, and fathers' and sons' scores on both the other-oriented and socially prescribed perfectionism subscales, did not correlate significantly. To examine whether this gender pattern concerning self-orientedperfectionism scores replicated, we computed correlations between students' self-rated, self-oriented perfectionism scores and students' ratings of parents' self-oriented perfectionism. Sons' self-ratingscorrelated more highly with their ratings of their fathers (r = .54,p < .001)than with their ratings of their mothers, r = .l 1, ns; t(55) = 2.86, p < .01. Furthermore, daughters' self-rated, self-oriented perfectionism scores correlated more highly with their ratings of their mothers (r = .51, p < .001) than with their ratings of their fathers, r = .27, p < .01; t(116) = 2.47, p < .05. The second prediction model we evaluated concerned the relation between parents' other-orientedperfectionism (i.e., the tendency to set exacting standards and to harshly evaluate others) and students' socially prescribed perfectionism (i.e., the tendency to view others' standards as too harsh and demanding). Table 6 presents the regression results for these analyses. A hierarchical regression was conducted, and the following predictors were entered into the equation at each step: student sex (Step I), mothers' and fathers' other-oriented perfectionism scores (Step 2; entered simultaneously), and finally, interactions between parents' other-oriented perfectionism scores and student biological sex (Step 3; entered simultaneously). Once again, all predictor variables were centered prior to comput-

TABLE 6 Hierarchical Multiple-Regression Results: Predicting Students' Socially Prescribed Perfectionismfrom Parents' Other-Oriented Perfectionism Scores Order ($Entry ofPredictor Variubles

Step 1 : Students' Sex Step 2: Mothers' OOP and Fathers' OOP Step 3: Interactions: Students' Sex x Mothers' SOP Students' Sex x Fathers' SOP

Model df

R2 Total

A R2

AF

F Overall

1,175 3, 173 5, 171

.04 .06 .07

.04 .02

7.87** 1.49 1.14

7.87** 3.63* 2.64*

.01

Note. OOP = other-oriented perfectionism; SOP = self-oriented perfectionism. *p < .05. **p < .01.

FAMILY PAlTERNS OF PERFECTIONISM

61

ing cross-product terms and prior to the variables being entered into the model. Table 6 indicates that parents' other-oriented perfectionism scores were not significantly related to students' socially prescribed perfectionism scores. Further, there was no evidence supporting moderator effects.


DISCUSSION This study investigated several major issues concerning the construct of perfectionism: (a) How much agreement is there between self-reports of perfectionism and observers' ratings on these same dimensions? (b) what are the patterns of self-oriented perfectionism among parents and their college-agechildren? and (c) what are the relations between parents' other-oriented perfectionism and students' socially prescribed perfectionism?

Convergence of Self- and Observer Ratings Previously, Hewitt and Flett (1991) reported evidence supporting the convergent validity of MPS scores obtained from 25 undergraduates and from 21 psychiatric outpatients when compared to observer ratings. Although some degree of convergence was obtained in our study, this varied as a function of who was being rated as well as the dimension of perfectionism under consideration. In general, more convergence was observed between daughters' self-reports and ratings from both parents. Further, daughters' ratings of parents converged more with parents' self-ratings than did sons' ratings. Thus, daughters' perfectionism appeared to be easier for parents to assess, and daughters appeared to be more accurate raters of parents' perfectionism. One possible explanation of these findings is that daughters may spend more time with their parents than do sons. If true, daughters' perfectionistictendencies may be more observable to parents and, in addition, daughters may be better assessors of their own parents' perfectionism. Admittedly, this is quite speculative and a study documenting gender differences in the amount of time spent with parents is needed.

Family Patterns of Perfectionism To determine whether parents' self-reported, self-oriented perfectionism scores aided in the prediction of students' scores, hierarchical multiple regression analyses were conducted. Further, correlational analyses were performed to test two competing hypotheses regarding perfectionistic thinking among parents and their children. The first hypothesis, or the primary caregiver hypothesis, asserts that lev-


62

VIETH AND TRULL

els of perfectionism in children are more likely to resemble the levels of perfectionism in mothers. The second hypothesis, or the same-sex modeling hypothesis, asserts that levels of perfectionism in children are more likely to resemble the levels characteristic of parents of the same sex. The results of this study are consistent with the same-sex modeling hypothesis. Not only were there significant and positive correlationsbetween the self-rated perfectionism scores of mothers and daughters and the self-rated perfectionism scores of fathers and sons, but the self-rated perfectionism scores of opposite-sex parent-child dyads were not positively associated. In fact, the self-rated perfectionism scores of mothers and sons were negatively associated. This pattern is similar to that observed by Parker (cited in Blatt, 1995) in a study of academically gifted children. In that study, although aportion of the variance in children's perfectionism scores could be accounted for by the scores of either parent, children's scores were better predicted by the scores of the same-sex parent. Furthermore, the present results are also consistent with the findings of Frost et al. (1991), who used the FMPS (Frost et al., 1990) to examine the patterns of perfectionism among parents and their college-agedaughters. In both of the small studies presented by Frost et al. (1991), the investigators found that daughters' overall perfectionism scores correlated significantly with mothers' overall perfectionism scores, but failed to correlate with fathers' overall perfectionism scores. It is also noteworthy that higher levels of self-rated perfectionism were positively associated with students' ratings of same-sex parents. Sons' self-ratings of self-oriented perfectionism correlated more highly with their ratings of their fathers than with their ratings of their mothers. Likewise, daughters' self-ratings of self-oriented perfectionism correlated more highly with their ratings of their mothers than with their ratings of their fathers on the same trait. Unlike Flett et al. (1995), we did not find evidence supporting an association between parents' tendency to set exacting standards and harshly criticize others and students' tendency to view others as harsh and demanding. Two major differences between ours and Flett et al.'s study may be responsible, at least partially, for the conflicting results. First, unlike Flett et al., we had actual reports from each parent versus relying on students' reports of parents' perfectionistic style. Second, students in Flett et al.'s study completed a measure of parental authority style for their parents, whereas we had parents self-rate their perfectionistic tendencies using the MPS. Clearly, more research is needed in this area to better assess the influence of parental authority style or perfectionism on the development of perfectionism styles in offspring.

Explaining mother-son and father-daughter associations. The relations observed between perfectionism levels in opposite-sex parent-child dyads warrants separate consideration. Whereas fathers' and daughters' self-oriented


FAMILY PAITERNS OF PERFECTIONISM

63

perfectionism scores were not correlated, mothers' and sons' scores were negatively correlated. In addition, whereas daughters' self-oriented perfectionism scores were correlated with their ratings of their mothers' and fathers' perfectionism, sons' perfectionism scores were uncorrelated with their ratings of mothers' perfectionism. In his article comparing the psychodynamic bases of normal and neurotic perfectionism, Hamachek (1978) introduced the concepts of positive modeling and negative modeling as antecedents of normal perfectionism. Positive modeling is defined as striving to emulate a model, whereas negative modeling is defined as striving to obtain qualities opposite those possessed by a model. Based on the patterns of perfectionism found in this study, we might hypothesize that daughters' levels of perfectionism result from positive modeling of the mother (i.e., emulating maternal strivings or the lack thereof). On the other hand, sons' levels of perfectionism may result from positive modeling of the father combined with negative modeling of the mother (i.e., reacting to maternal levels of perfectionism). This formulation is consistent with both personality theory and research in the sex-role literature. First, several theories of personality both acknowledge and account for the fact that males and females identify with or pattern themselves after same-sex models and/or reject opposite-sex models. For instance, Freud (cited in Basow, 1992) surmised that children initially identify with their same-sex parent as a way of allaying the anxiety associated with the Oedipal or Electra conflict. On a behavioral level, the social-learning theories of Mischel, Bandura, and others (also cited in Basow, 1992) posit that children attend to and imitate same-sex models, and fail to imitate opposite-sex models, based on patterns of reinforcement: In general, they are reinforced for imitating "sex-appropriate" behaviors and either punished or not rewarded for engaging in "sexinappropriate" behaviors. Although there are theoretical bases for same-sex modeling in both males and female participants, research from the sex-role literature suggests that males accept same-sex roles, and reject opposite-sex roles, to a greater extent than females. In their 1974 meta-analysis of sex differences across a number of domains, Maccoby and Jacklin (cited in Basow, 1992) concluded that boys show more behavioral sex-typing than girls. Urberg (1982), studying the development of the concepts of masculinity and femininity in a 3- to 7-year-old sample, arrived at the same conclusion. Based on their reviews of the literature, other authors (e.g., Serbin, Powlishta, & Gulko, 1993) have concluded that sex-typing in boys follows a linear trend with age (increasing or remaining stable through adolescence), in contrast to the curvilinear trend observed in girls. However, not only are boys more rigidly sex-typed than girls-meaning that they engage in more sex-appropriate behaviors-but they also appear to be less willing than girls to engage in opposite-sex behaviors. In their study of same-sex imitation, Bussey and Peny (1982) found that, although children of both sexes were equally accepting of same-sex be-


64

VIETH AND TRULL

havior, 8- to 11-year-old boys tended to reject opposite-sex behaviors more than girls in the same age range. Based on the aforementioned evidence, we would expect daughters in this study to resemble their fathers to a greater extent than sons resemble their mothers. At the very least, we might expect daughters to acknowledge andlor claim more similarity between themselves and their fathers than sons acknowledge between themselves and their mothers. Both of these expectations were borne out by the results presented here. In this study, whereas a significant negative association was found between sons' and mothers' self-rated, self-oriented perfectionism scores, daughters' and fathers' self-rated, self-oriented perfectionism scores were simply uncorrelated. In addition, whereas sons' self-ratings (self-oriented perfectionism) were not significantly associated with their ratings of their mothers, daughters' self-ratings were significantly correlated with the ratings they provided for their fathers.

Limitations and Recommendations for Future Study This study has a number of limitations, the most significant of which is the correlational nature of the data. Although perfectionism in college-age students appears to be related to the levels of perfectionism demonstrated by their same-sex parents, no assumptions may be made about the direction of causality. For instance, levels of perfectionism in parents may emerge in response to the perfectionistictendencies of their children, or perfectionistic tendencies in parents and children may develop simultaneously (Frost et a]., 1991). Further, this study assessed "perceptions" of perfectionism as opposed to including controlled observationsof perfectionisticbehaviors, and there may have been some bias (regarding perfectionistic tendencies) among those who completed and returned the questionnaires. Another limitation of this study is its cross-sectional design. It is possible that the patterns of perfectionism revealed by this study would not be replicated within different cohorts or within the same participants at different ages. Ideally, this study would be replicated using a sequential design, in which a number of cohorts are observed on more than one occasion, to control for both age and cohort effects. For example, Heatherton, Nichols, Mahamedi, and Keel (1995) found that levels of perfectionism were similar in two college cohorts separated by 10 years (i.e., 1982 and 1992). It is possible that the present findings hold for college-age individuals, but not for individuals in middle childhood or mid-life (age effect). Another consideration is that due to changes in sex roles over the last several years (Basow, 1992), children in each successive generation may resemble their same-sex parents to a lesser and lesser degree (i.e., a cohort effect). An additional limitation of this study is the homogeneous composition of the sample. By definition, all of the participants came from intact families, and all



65

were attending college. In addition, the overwhelming majority of participants were White and referred to themselves as either Catholic or Protestant. Finally, it is possible that perfectionism scores may be significantly related to whether partners remain married (or not), and we do not know whether the assumption that mothers are the primary caregivers holds true in perfectionistic families. The family patterns of perfectionism observed, therefore, may be unique to this demographically defined group. If possible, future investigations should enlarge the scope of study to examine the association between perfectionistic thinking in students and their primary caregivers, rather than in students and their biological parents. Over the last few decades, the increased divorce rate and the resulting changes in household composition (United States Bureau of the Census, 1993) require that studies of personality look beyond the biological parent-child relationship for influences on personality structure. Finally, subsequent studies of perfectionism might use behavioral genetics methods in order to examine genetic versus environmental contributions to perfectionism. For instance, concordance for perfectionism could be compared between monozygotic twins reared together and dizygotic twins reared together, or between monozygotic twins reared together and monozygotic twins reared apart. These approaches have been used to determine the heritabilities of a wide range of personality traits (e.g., Plomin, DeFries, & McClearn, 1990;Tellegen et al., 1988). Although this study focused primarily on the construct of self-oriented perfectionism, the findings may be relevant to other personality traits. It would be interesting to conduct similar studies of other well-defined traits in order to test the generalizability of the patterns revealed here. The resulting information would go far to advance our knowledge of the development of personality traits.

ACKNOWLEDGMENTS This article is based, in part, on Angela Z. Vieth's Master's thesis, which was completed at the University of Missouri under the direction of Timothy J. Trull. We thank Rebecca Perry, Julie Molitor, and Jennifer Weier, who assisted with the data management for this study. We also thank Arthur H. Goodwin for his valuable editorial assistance.

REFERENCES Aiken, L. S., &West, S. G. (1991).Multiple regression: Testingundinterpreting interactions. Newbury Park, CA: Sage. Barrow, J. C., &Moore, C. A. (1983). Group interventions with perfectionistic thinking. Personnel und Guidance Journal, 61, 61 2-61 5.


66

VIETH AND TRULL

Basow, S. A. (1992). Gender: Stereotypes and roles (3rd ed.). Pacific Grove, CA: BrooksICole. Belsky, J., Rovine, M., &Fish, M. (1989). The developing family system. In M. R. Gunnar & E. Thelen (Eds.), Systems and development: The Minnesota symposia on child psychology (Vol. 22, pp. 1 19-166). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Blankstein, K. R., Flett, G. L., Hewitt, P. L., & Eng, A. (1993). Dimensions of perfectionism ancl irrational fears: An examination with the Fear Survey Schedule. Personality and Individual Difirences, 15, 323-328. Blatt, S. J. (1995). The destructiveness of perfectionism: Implications for the treatment of depression. American Psychologist, 50, 1003-1020. Broday, S. F. (1988). Perfectionism and Millon basic personality patterns. Psychological Reports, 63, 791-794. Bums, D. D. (1980, November). The perfectionist's script for self-defeat. Psychology Today, pp. 34-52. Bussey, K., & Peny, D. G. (1982). Same-sex imitation: The avoidance of cross-sex models or tht: acceptance of same-sex models? Sex Roles, 8, 773-784. Carver, C. S., & Ganellen, R. J. (1983). Depression and components of self-punitiveness: High standards, self-criticism, and overgeneralization. Journal rfAbnorma1 Psychology, 92, 330-3137, Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Flett, G. L., Hewitt, P. L., Blankstein, K. R., & Dynin, C. (1994). Dimensions of perfectionism and Type A behavior. Personality and Individual Differences, 16, 477485. Flett, G. L., Hewitt, P. L., Blankstein, K. R., & Koledin, S. (1991). Dimensions of perfectionism1and irrational thinking. Journal cfRationa1-Emotive Therapy, 9, 185-201. Flett, G. L., Hewitt, P. L., & Dyck, D. G. (1989). Self-oriented perfectionism, neuroticismand anxiety. Personality and Individual Diflerences, 10, 731-735. Flett, G. L., Hewitt, P. L., &Singer, A. (1995). Perfectionism and parental authority styles. Individual Psychology, 51, 5 M 0 . Frost, R. O., Lahart, C. M., & Rosenblate, R. (1991). The development of perfectionism: A s~tudyof daughters and their parents. Cognitive Therapy and Research, 15, 469489. Frost, R. 0..Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism. Cognitive Therapy and Research, 14,449-468. Hamachek, D. E. (1978). Psychodynamics of normal and neurotic perfectionism. Psychology: A Journu1 ofHumun Behavior, 15, 27-33. Heathert0n.T. F., Nichols, P.,Mahamedi,F., & Keel, P. (1995). Body weight, dieting, andeatingdisorder symptomsamongcollegestudents, 1982to 1992.American Journal~fPsychiaty,152,1623-1629. Hewitt, P. L., & Dyck, D. G. (1986). Perfectionism, stress, and vulnerability to depression. Cognitive Therapy and Research, 10, 137-142. Hewitt, P. L., & Flett, G. L. (1989). The Multidimensional Perfectionism Scale: Development and validation. Canadian Psychology, 30, 339. Hewitt, P. L., & Flett, G. L. (1991). Perfectionism in the self and social contexts: Conceptualization, assessment, and association with psychopathology. Journal ofPersona1ityand Social Psychology, 60, 456-470. Hewitt, P. L., & Flett, G. L. (1993). Dimensions of perfectionism, daily stress, and depression: A test of the specific vulnerability hypothesis. Journal cfAbnormal Psychology, 102, 58-65. Hewitt, P. L., Flett, G. L., & Turnbull-Donovan, W. (1992). Perfectionism and suicide potential. British Journal ofclinical Psychology, 31, 181-190. Hewitt, P. L., Flett, G. L., Turnbull-Donovan, W., & Mikail, S. F. (1991). The Multidimensional Perfectionism Scale: Reliability, validity, and psychometric properties in psychiatric samples. Psycklological Assessment: A Journal c?fConsultingand Clinical Psychology, 3, 464-468. Hewitt, P. L., Flett, G. L., & Weber, C. (1994). Dimensions of perfectionism and suicide ideation. Cognitive Therapy and Research, 18,439-460.



67

Hollender, M. H. (1965). Perfectionism. Comprehensive Psychiatry, 6, 94-103. Kanfer, R., & Zeiss, A. M. (1983). Depression, interpersonal standard setting, and judgments of self-efficacy. Journal cfAbnormu1 Psychology, 92, 319-329. Kenny, D. A. (1994). Interpersonal perception: A social relations analysis. New York: Guilford. Lundh, L.-G., Broman, J.-E., Hetta, J., & Saboonchi, F. (1994). Perfectionism and insomnia. Scandinavian Journal o$Behuviour Therapy, 23, 3-18. Missildine, W. H. (1963). Your inner child ofthe past. New York: Simon & Schuster. Ordman, A. M., & Kirschenbaum, D. S. (1986). Bulimia: Assessment of eating, psychological adjustment, and familial characteristics. International h~urnai($Eating Disorders, 5, 865-878. Pacht, A. R. (1984). Reflections on perfection. American Psychologist, 39, 386-390. Parke, R. D. (1995). Fathers and families. In M. H. Borstein (Ed.), Handbookofbarenting: Vol. 3. Status and social conditions ($parenting (pp. 2 7 4 3 ) . Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Pirot, M. (1986). The pathological thought and dynamics of the perfectionist. Individual Psychology, 42, 51-58. Plomin, R., DeFries, J. C., & McClearn, G. E. (1990). Behavioral genetics: A primer (2nd ed.). New York: Freeman. Quadland, M. C. (1980). Private self-consciousness, attribution of responsibility, and perfectionistic thinking in secondary erectile dysfunction. Journal ($Sex & Marital Therapy, 6, 47-55. Rehm, L. P. (1977). A self-control model of depression. Behavior Therapy, 8, 787-804. Serbin, L. A., Powlishta, K.K., & Gulko, J. (1993). The development of sex typing in children. Monographs uf'theSociety for Research in Child Development, 58 (2, Serial No. 232). Slade, P. D. (1982). Towards a functional analysis of anorexia nervosa and bulimia nervosa. British Journal of Clinical Psychology, 21, 167-179. Slade, P. D., Newton, T., Butler, N. M., & Murphy, P. (199 I). Anexperimental analysis of perfectionism and dissatisfaction. British Journal ofClinica1 Psycholc~gy,30, 169-176. Sorotzkin, B. (1985). The quest for perfection: Avoiding guilt or avoiding shame. Psychotherapy, 22, 564-571. Tellegen, A,, Lykken, D. T., Bouchard, T. J., Wilcox, K. J., Segal, N. L., & Rich, S. (1988). Personality similarity in twins reared apart and together. Journal c$ Personality and Social Psychology, 54, 1031-1039. United States Bureau of the Census. (1993). Statistical abstract ofthe United States (1 13th ed.). Washington, DC: Reference Press. Urberg, K. A. (1982). The development of the concepts of masculinity and femininity in young children. Sex Roles, 8,659468. Waller, G., Wood, A,, Miller, J., & Slade, P. (1992). The development of neurotic perfectionism: A risk factor for unhealthy eating attitudes. British Review ($Bulimia and Anorexia Nervosa, 6, 57-62. Whiting, B. B., &Edwards, C. P. (1988). Children of diflerent worlds: Thefi)rmution ($social behavior. Cambridge, MA: Harvard University Press.

Timothy J. Trull Department of Psychology University of Missouri-Columbia 2 10 McAlester Hall Columbia, MO 6521 1 Received November 10, 1997 Revised August 29, 1998