Perfectionism, Maladaptive Self-presentation and …undergraduatelibrary.org/system/files/3789h.pdfPerfectionism, Maladaptive Self-presentation and Suicide Ideation in Male Retirees

Perfectionism, Maladaptive Self-presentation and Suicide Ideation in Male Retirees

Perfectionism, Maladaptive Self-presentation and Suicide Ideation in Male Retirees 1

Abstract

Perfectionism is a multifaceted trait that can be conceptualized by three distinguished dimensions:

Self-oriented Perfectionism, Other-oriented perfectionism and Socially Prescribed Perfectionism.

Perfectionists engage in maladaptive self-presentation styles, such as self-promotion, nondisplay and

nondisclosure to maintain their appeared perfection. Previous research has established the association

between trait perfectionism and suicide behaviors. I would suggest that perfectionistic male retirees would

be particularly vulnerable to suicide ideation due to stress brought by retirement and a lack of social

support caused by perfectionists’ nondisclosure tendencies. I investigated whether and how trait and self-

presentational components of perfectionism predict suicide ideation following retirement. A total of 92

mentally and physically healthy male retirees completed measures of perfectionism, self-worth and

suicide ideation at Time 1 before their retirement, and measures of nondisclosure, relationship specific

worry, perceived stress and suicide ideation at Time 2 five months later. The results showed that socially

prescribed perfectionism and all three perfectionistic self-presentational components were strongly

correlated with depression, hopelessness, subjective sense of alienation and perceived stress. Mediation

analysis suggested that the socially prescribed perfectionism indirectly led to suicide ideation through

nondisclosure then subjective social disconnection, d = .002 with CI.95 = [.0004, .01]. A moderated

mediation model revealed that the indirect pathway of suicide ideation was enhanced by contingence self-

worth from others’ approval and subjective stress, d = .0029 with CI.95 = [.0001, .01]. That is, for retirees

who felt that others demanded perfection from them, the more they based their self-worth on others’

approval, the more they tended to engage in nondisclosure, which led to subjective sense of alienation.

Consequentially, subjectively disconnected retirees with higher levels of stress are more likely to develop

suicide ideation compared to retirees with lower levels of stress. Interventions that promote help seeking

behaviors could be critical in suicide prevention among male retirees.


Suicide is a significant problem in the United States. Heron et al. (2008) reported that

suicide was the 11th

leading cause of death in the United States in 2006, accounting for 33,300

deaths. The suicide rate among male seniors are exceptionally high: black men experience the

second suicide peak at old age whereas suicide rate of white men peaks at the oldest group over

45 per 100,000 per year, almost four times the nation's overall age adjusted rate of 11.5 per

100,000 per year (Conwell, Orden, & Caine, 2011). Case controlled studies have established

several significant suicide risk factors such as physical illness (Juurlink, Herrmann, Szalai, Kopp,

& Redelmeier, 2004), cognitive deficit (Dombrovski et al., 2008), mental illness (Waern,

Rubenowitz, & Wilhelmson, 2003), and especially mood disorder (Conwell et al., 2009).

Because of the somewhat slow progress in this area, researchers are urged to go beyond medical

models and consider integrative models that address interpersonal factors (Prinstein, 2008).

Past and present theorists have postulated the disruption in interpersonal relationship as a

suicide risk factor (Baumeister & Leary, 1995; Joiner, 2005; Van Orden et al., 2010). Studies

using psychological autopsy have demonstrated that impaired interpersonal relationship often

precedes suicide (e.g., Miller, 1977). Case controlled studies also provide support for the

relationship between impairment in interpersonal relationship and suicide behavior (Duberstein,

Conwell, Conner, Eberly, Evinger, & Caine, 2004; Rubenowitz, Waern, Wilhelmson, &

Allebeck, 2001). There is also evidence suggesting that a lack of interpersonal interaction

increases suicide risk beyond physical and mental risk factors (Beautrais, 2002).

Perfectionism has been linked to suicide because of its interpersonal nature. Baumeister

(1990) proposed that suicide is sought as a mean to escape overwhelming distress caused by a

perfectionist’s inability to fulfill extremely high standard that others seemingly demand. Hewitt

and Flett (1993a; 2002) proposed that perfectionism could create, enhance and prolong


interpersonal stressors that lead to severe maladjustment such as suicide potential. Most recently,

Hewitt and colleagues (2006) have established perfectionism as interpersonal risk factor of

suicide in a model that incorporates the interpersonal components of perfectionism, impairment

of social relationship and suicide behavior, known as the Perfectionism Social Disconnection

Model (PSDM).

Perfectionism

There is general consensus that perfectionism is a multidimensional personality construct

and several models have been proposed to capture perfectionism (e.g., Frost, Marten,

Lahart, & Rosenblate, 1990; Hewitt & Flett, 1991b; Slaney, Rice, Mobley, Trippi, & Ashby,

2001). Our model emphasises on the interpersonal underpinning of perfectionism and we

describe perfectionism as dispositional traits (Hewitt & Flett, 1991b), interpersonal

presentational styles (Hewitt et al., 2003) and cognitive components (Flett, Hewitt, Blankstein, &

Gray, 1998). Perfectionism traits reflect the need to be perfect and consist of self-oriented

perfectionism (SOP, i.e., demand perfection of self), other-oriented perfectionism (OOP, i.e.,

demand perfection of others), and socially prescribed perfectionism (SPP, i.e., perceive that

others demand perfection of oneself). Since perfection cannot be achieved, perfectionists tend to

engage in Perfectionistic Self Presentation (PSP) in an attempt to appear perfect. PSP includes

three distinct interpersonal facets: perfectionistic self-promotion (i.e., actively promote one’s

supposed perfection), nondisplay of imperfection (i.e., avoid displaying one’s imperfection), and

nondisclosure of imperfection (i.e., avoid disclosing one’s imperfection; Hewitt et al., 2003).

Studies have shown that the trait perfectionism and self-presentational styles are

independent and differentially associated with various kinds of distress and disorders such as low

level of social self-esteem (Flett, Hewitt, & De Rosa, 1996), social anxiety (Nepon, Flett, Hewitt,


& Molnar, 2011), depressive symptoms (Sherry, Law, Hewitt, Flett, & Besser 2008), and eating

disorder symptoms (McGee, Hewitt, Sherry, Parkin, & Flett, 2005). The most severe outcome of

perfectionism is suicide and literature has shown strong association between SPP, but not SOP or

OOP, and suicide outcomes (O’Connor, 2007). For example, there is strong correlational

evidence that SPP is associated with suicide ideation (Dean & Range, 1996; Hewitt, Flett, &

Weber 1994; Hewitt, Newton, Flett, & Callender, 1997). Similarly, Klibert, Langhinrichsen-

Rohling, & Saito (2005) showed that SPP was positively correlated with suicidality irrespective

of method of assessment. Furthermore, SPP predicted unique variance of suicide threat and

suicide intention over and beyond hopelessness and depression (Hewitt, Flett, &

Turnbulldonovan, 1992; Dean, Range, & Goggin, 1996). In terms of suicide attempts, Hunter

and O'Connor (2003) found that SPP predicted parasuicides above and beyond the effects of

hopelessness, depression and anxiety. Finally, adult inpatients with alcoholism that made serious

suicide attempt were marked with elevated levels of SPP and low levels of OOP (Hewitt, Norton,

Flett, Callender, & Cowan, 1998).

The Perfectionism Social Disconnection Model

Although the evidence of maladaptive nature of trait and self-presentation perfectionism

has been accumulating, a comprehensive theory is needed to account for mechanisms by which

trait perfectionism manifests itself into maladaptive outcomes. Hewitt and Flett (2002) proposed

that perfectionism could interact with environmental stressors and lead to psychological distress

through stress enhancement, stress anticipation, stress generation and stress perpetuation. These

ideas were extended into the Perfectionism Social Disconnection Model (PSDM, Hewitt et al.,

2006). According to the PSDM, individuals with excessive interpersonal perfectionism are

motivated by excessive needs to be connected and belong with others. However, socially


prescribed perfectionists exhibit interpersonal hostility (Habke & Flynn, 2002; Hewitt et al.,

2003) and this interpersonal aggressiveness creates difficulties in establishing and maintaining

social relationships and leads to objective social disconnection (i.e., the perfectionistic individual

is not liked or cared for by others). In addition, socially prescribed perfectionists are

characterized by their interpersonal sensitivity (Hewitt & Flett, 1991). Because of their insecure

attachment (Chen et al., 2012), excessive neediness (Sherry et al., 2003) and sensitivity to

criticism; (Gilbert, Durrant, & McEwan, 2006), socially prescribed perfectionists could

misinterpret the environment and develop feelings of alienation and subjective social

disconnection (i.e., the perfectionistic individual feels rejected and excluded). As a result of both

objective and subjective interpersonal disconnection, perfectionistic individuals are likely to

experience psychological distress such as social anxiety, depressive symptoms, and suicidal

ideation (Baumeister & Leary, 1995).

There have not been many direct test of this relatively new model, though there is some

indirect evidence. For example, socially prescribed perfectionism and the three PSP facets, but

not self-oriented perfectionism, are associated with fear of negative evaluation (Flett et al., 1996),

excessive sensitivity to criticism (Gilbert, Durrant, & McEwan, 2006), and dependent attitudes,

such as need for admiration and desire to please others (Hewitt et al., 2003; Sherry, Hewitt, Flett,

& Harvey, 2003). In addition, there is evidence that interpersonal components of perfectionism

are associated with interpersonal problems and problematic relationships that affects social

connection. For example, Habke and Flynn (2002) showed that students with elevated SPP fall

in the hostile dominant quadrant of the interpersonal circumplex as do the PSP facets for men.

Women high in PSP tend to fall in the hostile submissive quadrant (Hewitt et al., 2001). Also,

Flett, Hewitt, Garshowitz, & Martin (1997) showed that SPP was associated with negative social


interactions including overt criticisms, lack of recognition and betrayal of others. More directly,

Hewitt et al. (2008) showed that clinicians who had conducted initial clinical interviews with

patients, disliked patients with higher levels of the PSP facets more than patients with lower

levels of PSP. Lastly, there is certainly long standing support for the link between social

disconnection and suicide tendency (see Heinrich & Gullone, 2006). For example, Joiner (2005)

proposed that social disconnection and a lack of belongingness are important in producing

suicidal behavior. Also, Levi-Bes et al., (2008) demonstrated that subjective social disconnection

caused by a lack of self-disclosure to intimate others is associated with compromised coping

capacity, and seriousness of the suicide attempts. Moreover, Levi-Bes et al., (in press) utilized

path analysis to show that subjective sense of social disconnection such as loneliness and sense

of alienation mediated the paths between both anxious and avoidant attachment patterns and

suicide lethality.

In a more direct test of the objective social disconnection part of the PSDM, Nepon et al.,

(2011) found that the relationship between nondisplay of imperfection and depressive symptoms

were mediated by negative social feedback suggesting that nondisplay of imperfection may have

led to objective social animosity which then produced depressive symptoms. Similarly, there is

evidence that although SPP and PSP facets are motivated by excessive need to belong and fear of

rejection, others seem to be able to see through the façade and as a result, dislike the

perfectionists (Hewitt et al., 2008). Moreover, Roxborough et al., (2012) showed that a marker of

objective social disconnection (i.e., being bullied) mediated the relationship between nondisplay

of imperfection and suicide potential. In terms of subjective social disconnection part of the

model, Sherry et al., (2008) showed that the relationship between SPP and depression symptoms

in a university student sample was mediated by perceived social support suggesting that


individuals high in SPP believed they received less social support resulting in increased

depression symptoms. Similarly, Roxborough et al., (2012) found in a sample of adolescent

psychiatric patients that a marker of subject social disconnection (i.e., social hopelessness)

mediated the relationship between all facets of PSP and suicide ideation suggesting that, in

adolescents, PSP may have led to subjective social disconnection which then produced suicide

ideation.

Present Study

The purpose of current study is two-fold: first, we wish to test and expand the validity of

the PSDM in relation to suicide ideation with a community sample consists of recent male

retirees. Second, we hope to take a closer look within the PSDM and study detailed pathways

through which SPP leads to suicide ideation. Most research of the PSDM utilized either college

students sample or psychiatric sample. Such samples are adequate in model testing but have

several drawbacks. For example, the effect size derived from student samples tend to differ from

non-student sample and the relatively low risk of the student sample could make the findings less

generalizable (Peterson, 2001). The psychiatric sample offers insights in the high risk population,

but relationship between suicide ideation and social factors cannot be studied in isolation since

mental illness is also a major compounding risk factor (Juurlink et al., 2004; Waern et al., 2003).

Such methodological difficulties could be resolved by using a sample of high-risk, non-

clinical sample. To study the relationship between perfectionism, social stressors and suicide

ideation, we recruited recent retirees because retirement process could be potentially stressful

(Solinge, 2008). We restricted participation to males due to their elevated suicide risk compared

to their female counterpart (Conwell et al., 2011; Coren & Hewitt, 1999). Retirees with other

major suicide risks, such as mental or physical illness, were excluded so the interpersonal factors


could be isolated and tested in relation with perfectionism and suicide ideation. This use of male

retiree sample not only could expand the validity and generalizability of the PSDM, but also has

potential to shed some light on the exceptional high suicide rate of male seniors. In order to test

the subjective portion of the PSDM, we utilized a two-wave design to assess the relationships

concurrently and longitudinally and the latter will allow detection of change in the outcomes as

well as an appropriate temporal test of the meditational model (Baron & Kenny, 1986). We

hypothesized that SPP at time 1 will lead to subjective feeling of social disconnection at time 2,

which then leads to concurrent suicide ideation (Figure 1).

The overall model places emphasis on individual trait (i.e., SPP) and cognitive

characteristic (i.e., interpersonal sensitivity) in leading to subjective social disconnection (Hewitt

et al., 2006). We wanted to take a closer look at the pathway between SPP and subjective social

disconnection because it is not entirely clear how individuals’ characteristics might lead to

subjective feeling of disconnection. In an effort to emphasize the interaction between individual

and his or her surrounding environment, we propose that PSP-nondisclosure, an interpersonal

behaviour style, as the mediating step between SPP and subjective social disconnection (Figure

2). By definition, failure to disclose to intimate others could be considered as a self-initiated

action of social disconnection between the individual and his or her environment. Such model

also enjoys support from literature. For example, there is a strong correlational evidence

suggesting SPP is associated with PSP-nondisclosure (Hewitt et al., 2003) and failure to disclose

is closely associated with social alienation and feeling of loneliness (Levi et al., in press). Chen

and her colleagues (2012) directly showed that nondisclosure mediated the relationship between

insecure attachment style, which is closely associated with SPP (Rice & Mirzadeh, 2000), and

subjective social disconnection.


Contingence self-worth from others’ approval was proposed as the moderator for the

relationship between SPP and nondisclosure. In the PSDM, both SPP and nondisclosure are

theorized to be motivated by powerful and excessive interpersonal need to be accepted and

approved of (Hewitt et al., 2006). Empirical study showed that dependent attitude, such as basing

self-worth on others’ approval, was positively correlated with SPP in psychiatric patients and

university students (Sherry, Hewitt, Flett, & Harvey, 2003). Since SPP is related to interpersonal

sensitivity such as heightened reaction to criticism (Gilbert et al., 2006) and evaluative concerns

perfectionism is strongly associated with contingent self-worth (Sturman, Flett, Hewitt, &

Rudolph, 2008), we speculated that contingent self-worth from others’ approval moderate the

relationship between SPP and nondisclosure (Figure 3, a path). That is, for perfectionists who

perceive others demand perfection of themselves, the more they based their self-worth on others’

opinion, the less they will disclose their shortcomings in an attempt to protect the sensitive and

fragile self.

Subjective stress was proposed as the moderator for the relationship between subjective

social disconnection and suicide ideation. Adjustment to the loss of the work role and the social

ties during retirement process is stressful (Solinge, 2008), and retirees who feel lonely and sense

of alienation could be especially vulnerable to these social stressors due to diminished social

support (Levi et al., 2008), passive coping strategies (Cacioppo et al., 2000a) and heightened

stress reactivity (Cacioppo et al., 2002a). Thus, stress could interact with subjective feelings of

social disconnection in leading to suicide ideation (Figure 3, b2 path). That is, subjectively

social-disconnected people who experience high levels of stress are more likely to develop

suicide ideation compared to the same people with low levels of stress.


To recap, the theoretical model predicts that:

1) Subjective social disconnection mediates the relationship between SPP at time1 and

suicide ideation at time2 (Figure1).

2) The relationship between SPP and suicide ideation is theorized to be mediated through

nondisclosure then subjective social disconnection (Figure 2).

3) The relationship between socially prescribed perfectionism and nondisclosure is

hypothesized to depend on contingence of self-worth from others’ approval whereas the

relationship between subjective social disconnection and suicide ideation is hypothesized to

depend on the level of perceived stress (Figure 3). Thus the indirect effect of SPP on suicide

ideation is conditional on the level of contingence self-worth and perceived stress.

Method

Participants

The initial sample was composed of 90 male retirees recruited from metro Vancouver

area. The age of the participants ranged from 43 to 80 (M = 59.65, SD = 5.17). The sample was

composed of 71Whites (78.8%), 6 Asians (6.6%), and 13 people (14.4%) who did not specify

their ethnicity. Seventy-two (82.2%) participants were married; five (5.6%) were single; seven

(7.8%) were separated or divorced; two (2.2%) were widowed, and two (2.2%) did not report

their marital status. The highest education levels of participants are high school (14%),

technology school (17%), university (45%), and graduate or professional school (24%). The

majority of the participants reported voluntary retirement (73%). There were no significant

differences between retirees who voluntarily retired and retirees who were forced into retirement

in terms of age, education level, perfectionism, and suicide ideation.


Measures

Multidimensional Perfectionism Scale (MPS; Hewitt & Flett, 1991b). The MPS is a 45-

item instrument that contains three subscales that measure self-oriented, other-oriented, and

socially prescribed perfectionism. Each subscale contains 15 items, which are scored on a 7-

point Likert scale ranging from 1 “strongly disagree” to 7 “strongly agree”. The scores of

individual subscale range from 15 to 105 and high scores indicate elevated perfectionism. The

reliability, validity and multidimensionality of the MPS have been demonstrated in both clinical

and non-clinical samples (Enns & Cox, 2002; Hewitt & Flett, 2004).

Perfectionistic Self-Presentation Scale (PSPS; Hewitt et al., 2003). The PSPS is a 27-

item instrument that contains three subscales measuring self-promotion, nondisplay of

imperfection, and nondisclosure of imperfection. The self-promotion and nondisplay subscale

each contains 10 items and the nondisclosure subscale contains 7 items. All items are scored on

a 7-point Likert scale ranging from 1 “strongly disagree” to 7 “strongly agree”. The scores of

self-promotion and nondisplay subscale range from 10 to 70 and the score of nondisclosure

subscale range from 7 to 49. In general, high scores indicate elevated levels of perfectionistic

self-presentation. The reliability and validity of PSPS has been documented (Hewitt et al., 2003).

Worry Domains Questionnaire (WDQ; Tallis, Eysenck, & Mathews, 1992). The WDQ is

a 25-item scale that measures domain specific worries (e.g. relationship, work incompetence, and

lack of confidence etc.).The relationship sub-scale is used as an indicator for subjective social

disconnection since the subscale taps into feelings of loneliness and subjective worry about

interpersonal relationship. The scale was designed for use on a nonclinical adult population and

the reliability and validity of the subscale has been documented (Rijsoort, Emmelkamp, &

Vervaeke, 1999).


Scale for Suicide Ideation (SSI; Beck, Steer, & Ranieri, 1988). The SSI is a 19-item scale

that measures themes such as wish to die, thought of active and passive attempt, frequency and

attitude toward ideation, capacity and planning, and final acts. Total score is the sum of all 19

items and can range from 0 (low ideation) to 38 (high ideation). SSI is a widely used suicide

measurement and several studies have documented its reliability and validity (Beck, Kovacs, &

Weissman, 1979; Clum & Curtin, 1993).

Perceived Stress Scale (PSS; Cohen, Kamarck, & Mermelstein, 1983). The PSS is a 14-

item self-report measure of nonspecific perceived stress. All items are scored on a 5-point Likert

scale ranging from 0 “never” to 4 “very often”. Seven items are reverse scored and individual

items are summed to calculate final score with higher scores reflecting greater perceived stress.

PSS is a widely used measure and several studies have supported its reliability and validity

(Cohen et al, 1983; Hewitt, Flett, & Mosher, 1992).

Contingencies of Self-Worth (CSW; Crocker, Luhtanen, Cooper, & Bouvrette, 2003). The

CSW is a 35-item self-report measure that contains seven domain specific sub-scales such as

“others’ approval”, “appearance”, “competition”, “academic competence”, “family support”,

“virtue”, and “god’s love”. Others’ approval subscale is used to measure self-worth derived from

approval and acceptance of others. All five items within the subscale are scored on a 7-point

Likert scale ranging from 1 “strongly disagree” to 7 “strongly agree”. Three items are reverse

scored and individual items are summed to calculate final score with higher scores reflecting

greater contingency of self-worth based on others’ view of the self. The reliability and validity of

the subscale has been demonstrated (Crocker et al., 2003).


Procedure

Based on power analysis, 71 participants are needed to achieve power of 0.8 with 5%

type I error rate if bias corrected bootstrapping is used to carry out mediational analysis with

medium effect size for both a path and b path (Fritz & MacKinnon, 2007; Preacher & Hayes,

2008). Therefore, assuming it is possible to retain 70% of the sample recruited at Time 1, at least

101 participants are needed at Time 1 to achieve reasonable statistical power for the mediational

analysis with longitudinal data.

118 recently retired male retirees were initially recruited through newspaper

advertisement; Participants were mailed the MPS and SSI, and a consent form indicating that

retirees who retired due to physical or mental illness were not eligible for the study. A total of 94

packages were returned; 4 participants later decided not to participate, leaving a total of 90

participants. Five months later, the participants were sent the PSPS, WDQ, CSW, PSS and SSI.

A total of 75 packages were returned and four packages were dropped due to excessive missing

data, leaving a total of 71 participants for both time1 and time21. Participants who completed

both Time1 and Time2 packages received $15.00 honorarium for their participation.

Data Analysis Strategies

Missing data was handled by conservative list-wise deletion since the missing at random

(MAR) requirement for multiple imputations (MI) was not fully met (Sterne et. al., 2009). In

order to test the moderation effects of perceived stress and self-worth derived from others’

approval, regression analyses were used to calculate the standardized simple slope and t statistics.

Bootstrapping procedures were used to construct confidence intervals of standardized effect size

1 The 15 participants who didn’t complete Time 2 measures did not differ from the 71 participants who completed

the Time 2 measures in terms of demographic variables, perfectionism traits or suicide ideation.


since standardized regression coefficients are not normally distributed (Kelley, 2005).

Bootstrapping resamples the original data with replacement to create 19999 bootstrap samples;

statistics of interest were computed for each bootstrap sample and an empirical confidence

interval was constructed based on all available resamples. The ratio of the obtained indirect

effect to the maximum possible indirect effect was computed following procedure listed by

Preacher and Kelley (2011). This effect size was only calculated for simple mediation in Figure

1since no viable procedure is available for chain mediation or moderated mediation.

To examine the conditioned mediation model, we followed the path analysis procedures

outlined in Preacher, Rucker, and Hayes (2007; note that the present model conforms to Model 4

in Figure 2). This method was chosen in an attempt to estimate effect size of mediational

relationship beyond qualitative description (Rucker, Preacher, Tormala, & Petty, 2011). The

point estimates of conditional effect sizes were calculated by multiplying standardized

coefficients of both a path, b1path, and b2 path under all possible conditions. Confidence

intervals were constructed by bootstrapping with accelerated bias correction method (Cheung,

2009). Suicide ideation at Time1 was controlled for all analysis for suicide ideation at Time 2.

Age was statistically controlled in b path since longitudinal study with large sample size

suggested that age was related to suicide ideation (Kuo, Gallo, & Tien, 2001). R is used for all

statistical computations.

Results

Means, standard deviations, coefficients alpha, and zero-order correlations are presented

in Table 1. Means and standard deviations approximated past studies that utilized the MPS

(Hewitt & Flett, 1991b), PSPS (Hewitt et al., 2003), and CSW (Crocker et al., 2003). However,

the mean of PSS (M = 16.94, SD = 9.61) was lower than reported by Cohen et al. (1983) for two


college student samples (M = 23.18 and 23.67) and for a community sample (M = 25.0). The

mean SSI at Time 1 and Time 2 were 0 and 0.1 respectively, which were substantially lower than

other community sample (M = 2.43) consisted of older adults (e.g., Miller, Segal, & Coolidge,

2001). Coefficients alpha ranged from .53 to .93. Both SSI at Time 1 and Time 2 had lower

coefficients alpha than in previous research (Beck et al., 1988). Finally, after adjusting

significance level for multiple comparisons (Holm, 1979), SPP was significantly correlated with

nondisclosure, perceived stress, and relationship specific worry. In addition, Perceived stress and

relationship specific worry were significantly correlated with suicide ideation at Time 2 (r = .41,

p < .01; r = .50, p < .01 respectively).

Mediational Analyses

Consistent with our hypothesis, the relationship between SPP and suicide ideation was

mediated by relationship specific worry, standardized indirect effect size d = .01 with CI.95 =

[.002, .022]. In other words, one standard unit change in SPP led to an average .01 standard

deviation increase in suicide ideation through relationship specific worries. The ratio of the

obtained indirect effect to the maximum possible indirect effect was calculated, = .221 (Figure

4) with 95% CI [.092, .440].2 Since the confidence interval excluded .09, a medium indirect

effect size was achieved (Preacher & Kelley, 2011). In addition, the indirect effect of SPP on

suicide ideation through nondisclosure then relationship specific worry was significant, d = 0.002

with CI.95 = [.0004, .01]. That is, one standard unit change in SPP led to an average .002 standard

deviation increase in suicide ideation through nondisclosure then relationship specific worries.

Moderated Mediation Analyses

2 500 bootstrap re-samples were used to construct the confidence interval via percentile bootstrapping.


Consistent with the expectation, the relationship between socially prescribed

perfectionism and none-disclosure was significantly moderated by amount of self-worth derived

from others’ approval, interaction b* = .31 with CI.95 = [.07, .48], t(66) = 3.81, p < .001. For

participants one standard deviation above the mean of contingency of self-worth, there was a

strong relationship between socially prescribed perfectionism and nondisclosure, b* = .69 with

CI.95 = [.51, .89], t(66) = 5.82, p < .001. For participants on the mean level of contingency of

self-worth, there was a moderate relationship between socially prescribed perfectionism and

nondisclosure, b* = .38 with CI.95 = [.19, .60], t(66) = 3.90, p < .001. In contrast, for participants

one standard deviation below the mean of contingency of self-worth there was a non-significant

relationship between socially prescribed perfectionism and nondisclosure, b* = .07 with CI.95 = [-

.21, .50], t(66) = .54, p > .05. Figure 5 provides a graphical representation of this interaction.

The association between relationship specific worry and suicide ideation was

significantly moderated by perceived stress while controlling for SPP and nondisclosure,

interaction b* = .23 with CI.95 = [-.06, .91], t(64) = 2.36 p < .05. For participants one standard

deviation above the mean of perceived stress, there was a significant relationship between

relationship specific worry and suicide ideation, b* = .41, with CI.95 = [.01, 1.11], t(64) = 2.87, p

<.01. In contrast, for participants on average level of perceived stress there was a non-significant

relationship between nondisclosure and suicide ideation, b* = .18 with CI.95 = [-.06, .99], t(64) =

1.08, p > .05. For participants one standard deviation below the mean of perceived stress there

was a non-significant relationship between nondisclosure and suicide ideation, b* = -.05 with

CI.95 = [-.66, 1.14], t(64) = -.02, p > .05. Regression diagnosis discovered the existence of

moderate multicollinearity, 5 < , <10. This multicollinearity could explain the


discrepancy between t statistics and empirically constructed confidence intervals. Figure 6

provides a graphical representation of this interaction.

Since self-worth moderates the relationship between SPP and nondisclosure (path a in

Figure 3) and perceived stress moderates the association between relationship specific worry and

suicide ideation while controlling for SPP and nondisclosure (path b2 in Figure 3), the magnitude

of the indirect effect depends on the levels of both self-worth and perceived stress. For

participants one standard deviation above the mean level of contingency of self-worth and one

standard deviation above the mean level of perceived stress, the indirect effect of SPP on suicide

ideation through nondisclosure then relationship specific worry was significant, d = .003 with

CI.95 = [.001, .01]. For participants on the mean level of contingency of self-worth and one

standard deviation above the mean level of perceived stress, the indirect effect was also

significant, d =.0016 with CI.95 = [.001, .006]. In contrast, the indirect effects were not

significant in all other conditions (Table 2). In sum, indirect effect was only significant when

perceived stress is one standard deviation above the mean. As contingence self-worth from

others’ approval increases, the magnitude of the indirect effect increases significantly. None of

the other PSP facets mediated this relationship in all conditions.

Discussion

The results provided strong support for the Perfectionism Social Disconnection Model.

The standardized indirect effect size in the chain mediation model is relatively small but

significant. The small effect size has to be interpreted in the content of small sample size and

inherited multicollinearity caused by strongly correlated predictor variables. By employing path

analysis, we demonstrated that trait perfectionism indirectly led to suicide ideation through


perfectionistic self-presentation and subjective sense of social disconnection. That is, by

concealing shortcomings, perfectionists’ attempt to connect with others ironically leads to a

subjective sense of disconnection. This disconnection, especially when combined with external

stress, in turn leads to suicide ideation.

The retirees in our sample were mentally and physically healthy, and they scored low on

distress measures compared to other community sample. However, perfectionism elevated their

suicide ideation significantly in merely five months. The increased stress and distress level also

makes the retirement process challenging. If resource allows, interventions need to focus on the

cause of retirement maladjustment, namely trait perfectionism, instead of symptoms. The

community could also establish outreach programs to encourage help seeking behaviours among

stressed male retirees, especially retirees who are reluctant to seek help due to their

perfectionism tendencies.


References

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social

psychological research: Conceptual, strategic, and statistical considerations. Journal of

Personality and Social Psychology, 51(6), 1173–1182.

Beck, A. T., Kovacs, M., & Weissman, A. (1979). Assessment of suicide ideation: The scale for

suicide ideation. Journal of Consulting and Clinical Psychology, 47(2), 343-352.

Beck, A. T., Steer, R., & Ranieri, W. (1988). Scale for suicide ideation: Psychometric properties

of a self-report version. Journal of Clinical Psychology, 44(4), 499-505.

Cacioppo, J. T., Ernst, J. M., Burleson, M. H., McClintock, M. K., Malarkey, W. B., Hawkley, L.

C., … Berntson, G. G. (2000). Lonely traits and concomitant physiological processes:

The MacArthur social neuroscience studies. International Journal of Psychophysiology,

35(2-3), 143–154.

Cacioppo, J. T., Hawkley, L. C., Crawford, L. E., Ernst, J. M., Burleson, M. H., Kowalewski, R.

B., … Berntson, G. G. (2002). Loneliness and health: Potential mechanisms.

Psychosomatic medicine, 64(3), 407–417.

Chen, C., Hewitt, P. L., Flett, G. L., Cassels, T. G., Birch, S., & Blasberg, J. S. (2012). Insecure

attachment, perfectionistic self-presentation, and social disconnection in adolescents.

Personality and Individual Differences, 52(8), 936–941.

Cheung, M. W.-L. (2009). Comparison of methods for constructing confidence intervals of

standardized indirect effects. Behavior Research Methods, 41(2), 425–438.

Clum, G. A., & Curtin, L. (1993). Validity and reactivity of a system of self-monitoring suicide

ideation. Journal of Psychopathology and Behavioral Assessment, 15(4), 375-385.

Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress.


Journal of Health and Social Behavior 24(4), 385-396.

Conwell, Y., Duberstein, P. R., Hirsch, J. K., Conner, K. R., Eberly, S., & Caine, E. D. (2009)

Health status and suicide in the second half of life. International Journal of Geriatric

Psychiatry 25(4), 371–379.

Conwell, Y., Orden, K. V., & Caine, E. D. (2011). Suicide in older adults. Psychiatric Clinics of

North America 34(2), 451–468.

Coren, S., & Hewitt, P. (1999). Sex differences in elderly suicide rates: Some predictive factors.

Aging & Mental Health, 3(2), 112-118.

Crocker, J., Luhtanen, R. K., Cooper, M. L., & Bouvrette, A. (2003). Contingencies of self-worth

in college students: Theory and measurement. Journal of Personality and Social

Psychology, 85(5), 894-908.

Dean, P. J., & Range, L. M. (1996). The escape theory of suicide and perfectionism in college

students. Death Studies, 20, 415–424.

Dean, P. J., Range, L. M., & Goggin, W. C. (1996). The escape theory of suicide in college

students: Testing a model that includes perfectionism. Suicide and Life-Threatening

Behavior, 26, 181–186.

Dombrovski, A. Y. et al. (2008). Cognitive performance in suicidal depressed elderly:

Preliminary report. American Journal of Geriatric Psychiatry, 16, 109–115.

Duberstein, P. R., Conwell, Y., Conner, K. R., Eberly, S., Evinger, J. S., & Caine, E. D. (2004).

Poor social integration and suicide: Fact or artifact? A case-control study. Psychological

Medicine, 34(7), 1331–1337.


Enns, M. W., & Cox, B. J. (2002). The nature and assessment of perfectionism. In G. L. Flett &

P. L. Hewitt (Eds.), Perfectionism: Theory, research, and treatment (pp. 33–62).

Washington, DC: American Psychological Association. doi:10.1037/10458-002

Flett, G. L., & Hewitt, P. L. (2002). Perfectionism: Theory, research, and treatment.

Washington, DC: American Psychological Association.

Flett, G. L., Hewitt, P. L., Blankstein, K. R., & Gray, L. (1998). Psychological distress and the

frequency of perfectionistic thinking. Journal of personality and social psychology, 75(5),

1363–81.

Flett, G. L., Hewitt, P. L., & De Rosa, T. (1996). Dimensions of perfectionism, psychosocial

adjustment, and social skills. Personality and Individual Differences, 20(2), 143–150.

Fritz, M. S., & MacKinnon, D. P. (2007). Required sample size to detect the mediated effect.

Psychological science, 18(3), 233–239.

Gilbert, P., Durrant, R., & McEwan, K. (2006). Investigating relationships between

perfectionism, forms and functions of self-criticism, and sensitivity to put-down.

Personality and Individual Differences, 41(7), 1299–1308.

Habke, M., & Flynn, C. (2002). Interpersonal aspects of trait perfectionism. In G. L. Flett & P. L.

Hewitt (Eds.), Perfectionism: Theory, research, and practice (pp.151-180). Washington,

DC: American Psychological Association.

Heinrich, L. M., & Gullone, E. (2006). The clinical significance of loneliness: A literature

review. Clinical Psychology Review, 26, 695–718.


Hewitt, P. L., & Flett, G. L. (1991b). Perfectionism in the self and social contexts:

Conceptualization, assessment, and association with psychopathology. Journal of

personality and social psychology, 60(3), 456–70.

Hewitt, P. L., & Flett, G. L. (2004). Multidimensional Perfectionism Scale: Technical manual.

Toronto, ON: Multi-Health Systems Inc.

Hewitt, P. L., & Flett, G. L. (2002). Perfectionism and stress processes in psychopathology. In G.

L. Flett & P. L. Hewitt (Eds.), Perfectionism: Theory, research, and treatment (pp. 255-

284). Washington, DC: American Psychological Association.

Hewitt, P. L., Flett, G. L., & Mosher, S. W. (1992). The Perceived Stress Scale: Factor structure

and relation to depression symptoms in a psychiatric sample. Journal of Psychopathology

and Behavioral Assessment, 14(3), 247–257. doi:10.1007/BF00962631

Hewitt, P. L., Flett, G. L., Sherry, S. B., & Caelian, C. (2006). Trait perfectionism dimensions

and suicidal behavior. In T. E. ELLIS (Ed.), Cognition and suicide: Theory, research,

and therapy (pp. 215–235). Washington, DC: American Psychological Association.

Hewitt, P. L., Flett, G. L., Sherry, S. B., Habke, M., Parkin, M., Lam, R. W., . . . Stein, M. B.

(2003). The interpersonal expression of perfection: Perfectionistic self-presentation and

psychological distress. Journal of Personality and Social Psychology, 84(6), 1303–1325.

doi:10.1037/0022-3514.84.6.1303

Hewitt, P. L., Flett, G. L., & Turnbulldonovan, W. (1992). Perfectionism and suicide potential.

British Journal of Clinical Psychology, 31(2), 181–190.


Hewitt, P. L., Flett, G. L., & Weber, C. (1994). Dimensions of perfectionism and suicide

ideation. Cognitive Therapy & Research, 18(5), 439-460.

Hewitt, P. L., Newton, J., Flett, G. L., & Callender, L. (1997). Perfectionism and suicide ideation

in adolescent psychiatric patients. Journal of Abnormal Child Psychology, 25(2), 95-101.

Hewitt, P. L., Norton, G. R., Flett, G. L., Callender, L., & Cowan, T. (1998). Dimensions of

perfectionism, hopelessness, and attempted suicide in a sample of alcoholics. Suicide and

Life-Threatening Behavior, 28, 395-406.

Holm, S. (1979). A simple sequentially rejective multiple test procedure. Scandinavian journal

of statistics, 6(2), 65–70.

Hunter, E., & O'Connor, R. (2003). Hopelessness and future thinking in parasuicide: The role of

perfectionism. The British Journal of Clinical Psychology / The British Psychological

Society, 42(4), 355-365.

Juurlink, D. N., Herrmann, N., Szalai, J. P., Kopp, A., & Redelmeier, D. A. (2004). Medical

illness and the risk of suicide in the elderly. Archives of Internal Medicine, 164(11),

1179–1184.

Kelley, K. (2005). The effects of nonnormal distributions on confidence intervals around the

standardized mean difference: Bootstrap and parametric confidence intervals.

Educational and Psychological Measurement, 65(1), 51–69.


Klibert, J. J., Langhinrichsen-Rohling, J., & Saito, M. (2005). Adaptive and maladaptive aspects

of self-oriented versus socially prescribed perfectionism. Journal of College Student

Development, 46(2), 141-156.

Kuo, W.-H., Gallo, J. J., & Tien, a. Y. (2001). Incidence of suicide ideation and attempts in

adults: the 13-year follow-up of a community sample in Baltimore, Maryland.

Psychological Medicine, 31(07).

McGee, B. J., Hewitt, P. L., Sherry, S. B., Parkin, M., & Flett, G. L. (2005). Perfectionistic self-

presentation, body image, and eating disorder symptoms. Body Image, 2(1), 29–40.

Miller, M. (1977). A psychological autopsy of a geriatric suicide. Journal of Geriatric

Psychiatry, 10(2):229– 242.

Miller, J. S., Segal, D. L., & Coolidge, F. L. (2001). A comparison of suicidal thinking and

reasons for living among younger and older adults. Death Studies, 25(4), 357–65.

O’Connor, R. C. (2007). The relations between perfectionism and suicidality: A systematic

review. Suicide and Life-Threatening Behavior, 37, 698–714.

Peterson, R. A. (2001). On the use of college students in social science research: Insights from a

second-order meta-analysis. Journal of Consumer Research, 28(3), 450–461.

Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and

comparing indirect effects in multiple mediator models. Behavior Research Methods,

40(3), 879–891.

Preacher, K. J., & Kelley, K. (2011). Effect size measures for mediation models: Quantitative

strategies for communicating indirect effects. Psychological Methods, 16(2), 93–115.


Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation

hypothesis: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42(1),

185-227.

Prinstein, M. J. (2008). Introduction to the special section on suicide and nonsuicidal self-injury:

a review of unique challenges and important directions for self-injury science. Journal of

Consulting and Clinical Psychology, 76(1), 1–8.

Rice, K. G., & Mirzadeh, S. A. (2000). Perfectionism, attachment, and adjustment. Journal of

Counselling Psychology, 47, 238–250.

Rijsoort, S. Van, Emmelkamp, P., & Vervaeke, G. (1999). The Penn state worry questionnaire

and the worry domains questionnaire: Structure, reliability and validity. Clinical

Psychology and Psychotherapy, 6(4), 297–307.

Rubenowitz, E., Waern, M., Wilhelmson, K., & Allebeck, P. (2001). Life events and

psychosocial factors in elderly suicides – a case–control study. Psychological Medicine,

31(07).

Rucker, D. D., Preacher, K. J., Tormala, Z. L., & Petty, R. E. (2011). Mediation analysis in

social psychology: Current practices and new recommendations. Social and Personality

Psychology Compass, 5(6), 359–371.

Slaney, R. B., Rice, K. G., Mobley, M., Trippi, J., & Ashby, J. S. (2001). The revised Almost

Perfect Scale. Measurement and Evaluation in Counseling and Development, 34, 130–

145.


Sherry, S. B., Hewitt, P. L., Flett, G. L., & Harvey, M. (2003). Perfectionism dimensions,

perfectionistic attitudes, dependent attitudes, and depression in psychiatric patients and

university students. Journal of Counseling Psychology, 50(3), 373–386.

Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equations

models. In S. Leinhart (Ed.), Sociological Methodology (pp. 290-312). San

Francisco: Jossey-Bass.

Sterne, J. a C., White, I. R., Carlin, J. B., Spratt, M., Royston, P., Kenward, M. G., Wood, a. M.,

et al. (2009). Multiple imputation for missing data in epidemiological and clinical

research: potential and pitfalls. Bmj, 338(jun29 1), b2393–b2393. doi:10.1136/bmj.b2393

Sturman, E. D., Flett, G. L., Hewitt, P. L., & Rudolph, S. G. (2008). Dimensions of

perfectionism and self-worth contingencies in depression. Journal of Rational-Emotive

& Cognitive-Behavior Therapy, 27(4), 213–231.

Tallis, F., Eysenck, M., & Mathews, A. (1992). A questionnaire for the measurement of

nonpathological worry. Personality and Individual Differences, 13(2), 161–168.

Van Orden, K. A., Witte, T. K., Cukrowicz, K. C., Braithwaite, S. R., Selby, E. A., & Joiner, T. r.

(2010). The interpersonal theory of suicide. Psychological Review, 117(2), 575-600.

doi:10.1037/a0018697

Van Solinge, H., & Henkens, K. (2008). Adjustment to and satisfaction with retirement: two of a

kind? Psychology and Aging, 23(2), 422–34.


Table 1

Means, Standard Deviations, Coefficients Alpha, and Zero-Order Correlations

Variable SOP OOP SPP SP DSPY DISC WDQ CSW PSS SSI1 SSI2

SOP -- 0.77** 0.69** 0.67** 0.48** 0.50** 0.30 0.19 0.26 -- 0.16

OOP -- 0.54** 0.55** 0.38* 0.38* 0.31 0.19 0.24 -- 0.26

SPP -- 0.52** 0.52** 0.48** 0.44** 0.20 0.41* -- 0.18

SP -- 0.79** 0.79** 0.40* 0.35** 0.36* -- 0.18

DSPY -- 0.78** 0.50** 0.50** 0.59** -- 0.22

DISC -- 0.45** 0.29 0.49** -- 0.23

WDQ -- 0.26 0.63** -- 0.50**

CSW -- 0.35* -- 0.03

PSS -- -- 0.41**

SSI2 -- --

M 60.17 51.14 44.08 35.76 35.37 21.65 7.32 4.33 16.94 0 0.1

SD 16.99 11.75 12.88 12.96 12.19 7.64 3.09 1.18 9.61 0 0.45

α 0.91 0.80 0.85 0.91 0.89 0.84 0.85 0.74 0.93 N/A 0.56

Note. * p < .05. ** p < .01. Significance levels are adjusted for multiple testing (Holm, 1979).

SOP = self-oriented perfectionism; OOP = other-oriented perfectionism; SPP = socially

prescribed perfectionism; SP = self-promotion; DSPY = nondisplay; DISC = nondisclosure; SSI

=suicide ideation; WDQ= worry domain questionnaire, relationship; CSW= contingency of self-

worth, others’ approval; PSS = perceived stress. M = means; SD = standard deviations; α =

coefficients alpha.


Table 2

Moderated Mediation Analysis and Confidence Interval

CSW (a path) PSS (b2 path) Estimate Standard Error .95 CI

High High .0029 .0027 [.0001, .0104]*

Medium High .0013 .0016 [-.0003, .0084]

Low High .0003 .0030 [-.0061, .0066]

High Medium .0015 .0016 [.0001, .0060]*

Medium Medium .0007 .0009 [-.0002, .0048]

Low Medium .0002 .0018 [-.0036, .0037]

High Low .0003 .0011 [-.0009, .0041]

Medium Low .0001 .0005 [-.0002, .0029]

Low Low .0000 .0001 [-.0025, .0012]

Note. * Confidence interval does not cross zero. The estimates are standardized indirect effect

size. High = +1 SD, Medium = mean, Low= -1 SD.


Figure 1. Theoretical Model for Subjective Social Disconnection

Figure 2. Theoretical Model for Two-Step Mediation

Subjective Disconnection

Socially Prescribed Perfectionism

Suicide Ideation

Figure 1. Subjective Social Disconnection Mediating Relationship between Social

Prescribed Perfectionism and Suicide Ideation.

Subjective

Disconnection

Socially Prescribed

Perfectionism Suicide Ideation

Nondisclosure

Figure 2. The Chain Mediation Model Being Tested. The Link Between Socially Prescribed

Perfectionism and Suicide Ideation Is Mediated through Nondisclosure then Subjective Social

Disconnection.



Figure 4. Effect Size for Mediation Analysis

Figure 4. Ratio between achieved indirect effect and maximum indirect effect.

Graph generated by R syntax developed by Kelly 2011


Figure 5. Simple Slope Analysis for Path a

Figure 5. Simple Slopes for the Interaction between Socially Prescribed Perfectionism and

Contingence Self-Worth in Predicting Nondisclosure.


Figure 6. Simple Slope Analysis for Path b2

Figure 6. Simple Slopes for the Interaction between Relationship Specific Worry and Perceived

Stress in Predicting Suicide Ideation.

Documents

Perfectionism, Maladaptive Self-presentation and …undergraduatelibrary.org/system/files/3789h.pdfPerfectionism, Maladaptive Self-presentation and Suicide Ideation in Male Retirees