The Effect of Perfectionism on Career Adaptability and Career Decision-Making Difficulties

Abstract

This study investigated categorization of perfectionism subtypes for Chinese undergraduates and the effects of perfectionism subtypes on career outcomes based on two prominent, competing models of perfectionism, the tripartite model and 2 × 2 model. Indices of career outcome were defined with career adaptability (positive) and career decision-making difficulties (negative). The results of both cluster analysis and latent profile analysis coincided with the four-subtype structure of the 2 × 2 model. The result of Bolck–Croon–Hagenaars modeling indicated that the pure high standard subtypes were the most functional while pure discrepancy subtypes were most dysfunctional. Mixed perfectionism subtypes were identified as having high career adaptability but also high risk for career decision-making while nonperfectionism subtypes possess low career decision-making difficulties but also low career adaptability. Based on these findings for perfectionism subtypes, we extrapolate practical recommendations for how this information could be pertinent to career counseling.

Keywords

perfectionism career adaptability career decision-making difficulties Chinese undergraduates latent profile analysis

Perfectionism is defined as “personality style characterized by striving for flawlessness and setting of excessively high standards for performance accompanied by tendencies for overly critical evaluations of one’s behavior” (Stoeber & Otto, 2006, p. 295). Perfectionism was identified as relevant to career development (e.g., Andrews, Bullock-Yowell, Dahlen, & Nicholson, 2014; Lehmann & Konstam, 2011; Page, Bruch, & Haase, 2008) and as a potential target for career counseling intervention. Most existing quantitative research about the interplay of perfectionism and career development focuses on dimensions of perfectionism used in regression or SEM models to predict career outcomes (e.g., Stoeber, Mutinelli, & Corr, 2016). Although these studies have deepened understanding about how perfectionism predicts career outcomes, results such as the interactive effects of perfectionism dimensions have yet to become relevant for practitioners, given the lack of practical information from these findings. Specifically, it is the perfectionism subtypes defined by within-person combinations of various perfectionism dimensions, rather than the dimensions themselves, that are identified as meaningful units to differentiate individual function, potential, and risk (Gaudreau & Thompson, 2010), although how these subtypes differentiate in career outcomes has yet to be elucidated for the benefit of counselors who could assess for subtype to fine-tune prevention strategies and targeted interventions in their work. Therefore, this study explores the relationships between perfectionism subtypes and career outcomes with emphasis on how this information could be pertinent in real-world context for at-risk individuals in schools.

Regarding subtypes of perfectionism, there are currently two prominent but competing theoretical models: the tripartite model (Stoeber & Otto, 2006) and the 2 × 2 model (Gaudreau & Thompson, 2010). Both models have demonstrated variability in compatibility for their definitions of perfectionism subtypes across various life domains such as school (Arana & Furlan, 2016), athletics (Gaudreau & Verner-Filion, 2012), and work (Li, Hou, Chi, Liu & Hager, 2014) due to inconsistent empirical evidence. Little research has been conducted on how these models fit within the domain of career development which is a novel consideration fulfilled through this study.

Two variables were chosen to operationalize career outcomes for this study based on the target population: career adaptability as positive index (Savickas, 1997, 2013) to capture the proactive mechanisms of undergraduates preparing to transition from school to work and emotional and personality-related career decision-making difficulties (EPCD) as negative index (Saka, Gati, & Kelly, 2008) to capture the uncertainty and analysis paralysis for many within this same domain. Career adaptability reflects the readiness to cope with career transition and future development tasks (Savickas, 1997, 2013), while EPCD reflects the various dysfunctional states of individuals in career decision-making (Gati, Krausz, & Osipow, 1996).

Overall, this study serves to elucidate the validity of perfectionism subtypes for either perfectionism model (tripartite and 2 × 2) within the context of career development for Chinese undergraduates and to investigate the predictive power of perfectionism subtypes for positive and negative career outcomes in accordance to either model. The findings of this study will be essential to guiding the practice and strategies of career counselors; by identifying how to assess for personality traits indicative of perfectionism subtype and differentiating between subtypes, practitioners can adapt targeted interventions for at-risk students to help them capitalize on their potential and mediate for positive career outcomes.

Perfectionism

Adaptive and maladaptive perfectionism

Perfectionism has been conceptualized as a multidimensional construct with a two-sided nature (Hamachek, 1978). Factor analysis based on various measures of perfectionism (Cox, Enns, & Clara, 2002) and literature review on empirical studies reviewing facets of perfectionism (Stoeber & Otto, 2006) bolsters the claim that perfectionism can be parsed into two basic dimensions of adaptive and maladaptive perfectionism.¹ Adaptive perfectionisms include emphasizing order and precision within self-directed establishment of high standards for achievement. Maladaptive perfectionism includes the concern with mistakes, evaluation, and comparison to the standards of others in conjunction with discrepancy between self-standards and actual performance (Cox et al., 2002; Stoeber & Otto, 2006).

However, for students, this categorization is rarely black-and-white as individuals can possess traits of both adaptive and maladaptive perfectionism (Hewitt & Flett, 1991; Gaudreau & Thompson, 2010; Stoeber & Otto, 2006). Within-person combination of these dimensions demonstrates a more accurate representation of individual manifestation for integral perfectionism (Gaudreau & Thompson, 2010; Stoeber & Otto, 2006). Still, controversy looms as to how these dimensions combine to form and classify subtypes of perfectionism and what differences exist between the subtypes to mediate positive or negative outcomes for individuals. Two prominent and competing models of perfectionism, the tripartite model (Stoeber, 2012; Stoeber & Otto, 2006) and the 2 × 2 model (Gaudreau, 2013; Gaudreau & Thompson, 2010), conflict with how perfectionism subtypes are classified and by extension how their subtypes are indicative of function.

The tripartite model and the 2 × 2 model

The tripartite model was first proposed by Stoeber and Otto (2006). The three subtypes they identified were healthy perfectionists (HPs; high striving and low concern), unhealthy perfectionists (UPs; high striving and high concern), and nonperfectionists (low striving). Adaptive perfectionism dimension differentiates perfectionists from nonperfectionists, while maladaptive concern differentiates between healthy and unhealthy forms of perfectionism. Valence of these subtypes was confirmed from review of empirical studies investigating association of perfectionism to psychological outcomes with previously confirmed positive and negative valences such as life satisfaction, positive and negative affect, internal and external locus of control, and attachment avoidance (Stoeber & Otto, 2006). HPs were demonstrated to have highest levels of positive psychological outcomes while UPs demonstrated highest levels of negative psychological outcomes. This tripartite model and associated valences have been verified in Canadian and Chinese samples (Smith, Saklofske, Yan, & Sherry, 2015).

Gaudreau and Thompson (2010; Gaudreau, 2013) proposed the 2 × 2 model of perfectionism, creating a four-subtype construct combining dimensions of adaptive personal standards perfectionism (PSP) and maladaptive evaluative concerns perfectionism (ECP). The four subtypes they identified were nonperfectionism (low PSP and low ECP), pure personal standards perfectionism (high PSP and low ECP), pure evaluative concerns perfectionism (low PSP and high ECP), and mixed perfectionism (MP; high PCP and high ECP). Gaudreau and Thompson (2010) distinguished the function and valence of these subtypes through four hypotheses evaluated and validated in research conducted at a Canadian university. Their model has also been supported from studies in China (Li et al., 2014). Three competing hypotheses were about distinction between pure personal standards perfectionism (pure PSP) and nonperfectionism (NP). Hypothesis 1a postulates that pure PSP is more functional than NP; Hypothesis 1b postulates pure PSP is unhealthier than NP; Hypothesis 1c postulates these two subtype do not differ in psychological function. Hypothesis 2 establishes pure evaluative concerns perfectionism (pure ECP) has the most negative outcomes. Hypothesis 3 establishes that MP is more functional than pure ECP. Hypothesis 4 establishes that MP is more dysfunctional than pure PSP (Gaudreau & Thompson, 2010).

Two primary distinctions between the two models are the source of conflict and controversy. First, the tripartite model regarded people with low striving/standards as nonperfectionism subtype while the 2 × 2 model divided the counterpart into two subtypes, nonperfectionism and pure maladaptive perfectionism based on the level of the maladaptive dimension, and distinguished them in function. Second, the tripartite model identified the UP subtype with both high adaptive and maladaptive dimensions as most detrimental while the same combination of dimensions were identified as MP subtype in the 2 × 2 model that is not the most detrimental to function.

Perfectionism and career outcomes

Since perfectionism is essentially related to self-standard setting and goal striving (Cox et al., 2002), it reflects three basic processes involved in career development: setting and striving for goals, constructing self-perception, and adapting for environmental needs (Lent, 2005; Savickas, 2013). Perfectionism are associated with various self-construction factors such as self-esteem (Mobley, Slaney, & Rice, 2005), locus of control (Periasamy & Ashby, 2002), and self-criticism (Grzegorek, Slaney, Franze, & Rice, 2004). Furthermore, capacity for adaptation to environmental demands is reflected in the fundamental two-dimension conceptualization of perfectionism, consistently comprised of an adaptive and maladaptive dimension. Overall, adaptive or healthy perfectionism (setting appropriate high self-standards) indicates well-functioning self-construction, setting and striving goals, and responding to the environment, while maladaptive or unhealthy perfectionism (setting inappropriate self-standards) indicates problems or obstacles in these processes of career development.

The three basic processes of career development (Lent, 2005; Savickas, 2013) can be evaluated by two indicators, positive career adaptability and negative career decision-making difficulties, and thus we regarded them as outcomes of perfectionism. Savickas (1997) defines career adaptability as the “readiness to cope with predictable tasks of preparing for and participating in the work role and with the unpredictable adjustments promoted by changes in work and working conditions” (p. 254). Career adaptability is a synthetic index reflecting the psychosocial resources of life span career development (Savickas, 2013) and is comprised of four dimensions: concern, control, curiosity, and confidence. Concern and curiosity reflect the planning attitude and initiative exploration in the process of career goal setting and striving; control and confidence reflect the career-related part of self-concept construction; the profile of these four dimensions reflects one’s function of adaptation to changing environment (Savickas, 2013). Career decision-making difficulties encapsulate various kinds of problems that can manifest before, during, or after the entire decision-making process (Gati et al., 1996), including the emotional and personality aspects of career decision-making difficulties proposed by Saka, Gati, and Kelly (2008). They identified three clusters of influence on career decision-making difficulties: pessimistic views, anxiety, and self-concept and identity. Pessimistic views refer to negative perceptions of self within the context of vocation. Anxiety refers to a negative emotional state occurring during the career goal setting process. Self-concept and identity refers to some problems about self-esteem, self-identity crystallization, and self-boundary. Thus, these three clusters reflect the maladaptive performances in goal setting and self-construction process of career development.

Currently, only limited and indirect empirical evidence supports the relationship between perfectionism and career adaptability as well as career decision-making difficulties. Adaptive perfectionism has been found to positively correlate with a constructive career planning attitude (Stoeber et al., 2016) and self-efficacy in career decision-making (Page et al., 2008) while also being negatively correlated with pessimistic career thoughts (Andrews et al., 2014). Maladaptive perfectionism has demonstrated opposite associations with career planning attitude, career decision-making, and career thoughts. These findings implicate a relationship between the dimensions of perfectionism and a healthy career development process involving career adaptability, although further studies are necessary to clarify the nature of this relationship. Similarly, the dimensions of perfectionism have been validated to predict career indecision and information-related career decision-making difficulties (Lehmann & Konstam, 2011; Leong & Chervinko, 1996; Page et al., 2008), although no study has directly related the personality and emotion-related career decision-making difficulties to perfectionism.

Objectives and hypotheses

The objectives of this study were to (a) determine which model of perfectionism would better represent perfectionism subtypes of Chinese undergraduates and (b) investigate the relationship between perfectionism subtypes and career development variables of career adaptability and career decision-making difficulties. Hypotheses were constructed for both the tripartite and 2 × 2 models of perfectionism (Figures 1 and 2) to consider all potential outcomes since no previous study has been conducted on the applicability of these models to Chinese undergraduates and career development.

Should the tripartite model have been validated, research participants would have been organized into clusters of HPs, UPs, and nonperfectionists (NPs) in accordance with the previously established subtypes. The HP cluster was postulated to demonstrate the highest career adaptability (Hypothesis A1a) and lowest career decision-making difficulty (Hypothesis A1b). The UP cluster was postulated to demonstrate the lowest career adaptability (Hypothesis A2a) and the highest career decision-making difficulty (Hypothesis A2b).

Should the 2 × 2 model have been validated, research participants would have been organized into clusters of nonperfectionism (NP), pure high standard (PHS), pure discrepancy (PD), and MP. The NP cluster was postulated to demonstrate higher career adaptability (Hypothesis B1a) and lower career decision-making difficulty (Hypothesis B1b) than PD. The MP cluster was also postulated to have higher career adaptability (Hypothesis B2a) and lower career decision-making difficulty (Hypothesis B2b) than PD. PHS was postulated to have higher career adaptability (Hypothesis B3a) and lower career decision-making difficulty (Hypothesis B3b) than MP. A hypothesis was not formed for the relationship between NP and PHS as the nature of such relationship has yet to be confirmed within the foundation of the 2 × 2 model so comparisons between the clusters was taken from the data in an exploratory manner.

Method

Participants and Procedure

A total of 750 questionnaires was delivered to undergraduates of five universities in China: two in the North region, one in the South region, one in the East region, and one in the West region. The students were voluntarily participated and were given a 5-Yuan gift of appreciation. Seven hundred and twelve questionnaires (94.93%) were returned. Initial analysis of the returned questionnaires resulted in exclusion of blank questionnaires and those with obvious patterns (i.e., 123454321) indicative of feigned participation. An additional 11 questionnaires were excluded when difference values between two valid test items were larger than four according to criteria for valid test items of the EPCD (Saka et al., 2008). Overall, 672 questionnaires (89.60%) were validated for further data analysis.

Within our sample of 672 undergraduates, 262 individuals were male (38.99%), 407 individuals were female (60.57%), and 3 individuals did not report gender (0.45%). The age of participants ranged from 21 to 27 years old and the average age was 23.30 years old (SD = 1.04). Because this study focused on the career development process, only sophomores (n = 334, 49.70%) and juniors (n = 332, 47.92%) were recruited to participate in the survey because of their greater likelihood to be active in career decision-making as compared to freshmen (not yet contemplating) and seniors (already decided) so the measurement of career decision-making difficulties would be as relevant as possible.

Measures

Perfectionism

The Chinese version of the Almost Perfect Scale–Revised (Slaney, Rice, Mobley, Trippi, & Ashby, 2001) was used to measure adaptive and maladaptive perfectionism. The original assessment consisted of three subscales: High Standard (7 items), Discrepancy (12 items), and Order (4 items). Because order is not a core dimension of perfectionism (Rice, Ashby, & Slaney, 2007) and theoretically has minimal association with career outcomes, the order subscale was omitted in this study. All items were on a 7-point Likert-type scale (1 = strongly disagree and 7 = strongly agree). Its internal consistency (.74 for high standard and .85 for discrepancy), test–retest reliability (.72 for high standard and .71 for discrepancy), and structure validity (root mean square error of approximation [RMSEA] = .077, comparative fit index [CFI] = .91, non-normal fit index [NNFI/TLI] = .90, standardized root mean square residual [SRMR] = .067) were good for Chinese population (Yang, Liang, Zhang, & Wu, 2007). In this study, Cronbach’s α was .77 for the high standard subscale and .86 for the discrepancy subscale.

Career adaptability

The Chinese version of the Career Adapt-Abilities Scale (CAAS; Hou, Leung, Li, Li, & Xu, 2012; Savickas & Porfeli, 2012) was used to measure the positive career development outcome of career adaptability. CAAS contains 24 items gauging four dimensional subscales: Concern, Control, Curiosity, and Confidence. All items were on a 5-point Likert-type scale (1 = not strong and 5 = strongest) related to self-perception of the abilities presented in each item. The Chinese version of the CAAS has good structure validity (RMSEA = .064, SRMR = .057) and internal consistency (.79, .64, .71, .74 and .89 for concern, control, curiosity, confidence, and total scale; Hou et al., 2012). In present study, Cronbach’s α was .92 for overall scale, .82 for concern, .77 for control, .74 for curiosity, and .81 for confidence.

Career decision-making difficulties

The Chinese version of the Emotional and Personality-Related Career Decision-Making Difficulties Scale–Short Form (EPCD-SF; Ikeda, Hou, Liu, Li, & Itamar, 2016; Saka et al., 2008) was used to measure the negative career development outcome of career decision-making difficulties. EPCD-SF contains 1 warm-up item, 2 validity-test items (e.g., “I am happy when something good happens to me”), and three dimensional subscales: pessimistic views (6 items), anxiety (8 items), and self-concept and identity (6 items). All items were on a 9-point Likert-type scale (1 = does not describe me at all and 9 = describes me well) while higher ratings represented a higher level of career decision-making difficulty. The Chinese Version of the EPCD-SF has good structure validity (RMSEA = .05, CFI = .93, TLI = .91, SRMR = .05), criterion-related validity, and internal consistency (.66, .78, .74, and .86 for pessimistic view, anxiety, self-concept and identity, and total scale, respectively; Ikeda et al., 2016). In this study, Cronbach’s α was .90 for total scale, .73 for pessimistic views, .86 for anxiety, and .81 for self-concept and identity.

Statistical Analysis

We used both traditional two-step cluster analysis procedure and a newly developed latent profile analysis (LPA) method to identify the subtypes of perfectionism, considering the comparability of the result to previous studies and the preciseness of perfectionism subtype classification. The traditional procedure (Rice & Ashby, 2007) involves a hierarchical to generate the average scores of the dimensions in each cluster (centroids), respectively, in three-cluster and four-cluster solution and a nonhierarchical k-means cluster analyses based on the centroids to get the results of three-cluster and four-cluster classification.

Then, the exploratory and confirmatory LPA were applied using MPlus 7.4 (Muthén & Muthén, 1998–2015) to detect the optimal model. The explorative LPA involved fitting several models with increasing number of classes. The confirmatory LPA allows us to fix the means of the two classification dimensions (i.e., high standard and discrepancy) using the final centroids generated from the traditional cluster analysis. Next, all these LPA models were compared and the optimal model was selected based on several criteria (Nylund, Asparouhov, & Muthén, 2007): Akaike information criterion (AIC), Bayesian information criterion (BIC), sample-size adjusted Bayesian information criterion (a-BIC), Lo–Mendell–Rubin likelihood ratio test (LRT), bootstrap likelihood ratio test (BLRT), and entropy. Lower values of AIC, BIC, and a-BIC indicate better balance between model fit and parsimony. The significant p values of LRT and BLRT means the current k-class model fits better than the model with k − 1 classes. Entropy is an index of classification quality that values from 0 to 1. Its higher value indicates better differentiation between profiles. In addition, the solutions with profile size less than 32 (< 5%) or with any profile that were theoretically meaningless were excluded regardless of the model fit.

After the optimal model was chosen, the Bolck–Croon–Hagenaars (BCH) approach (Asparouhov & Muthén, 2014; Bakk & Vermunt, 2016) was applied to examine the association between perfectionism and career-related outcomes. BCH is the currently most recommended method to investigate the differences among groups in continuous distal outcomes. Compared with traditional method to examine between-group difference like analysis of variance, BCH accounts for classification probability and increases the reliability of parameter estimation. BCH approach consists of three steps: conducing LPA, generating weight variables that reflecting the error of the latent class variable, and fixing this weights to the latent class variable meanwhile estimating the means of outcomes in each class. Then, Wald test (Muthén & Muthén, 2015) was used to examine the pairwise differences between any two classes on the outcome variables.

Results

Preliminary Analysis

First of all, missing value analysis was implemented and the result demonstrated that no variable or case had a missing value of more than 10% which was a common criterion for inclusion (Hair, Black, Babin, & Anderson, 2009) and thus no variable or case was deleted. Then, descriptive analysis and correlational analysis were executed and the results are presented in Table 1. The dimensions of perfectionism were most often associated with same valence indices of career outcomes.

Table 1.

Means, Standard Deviation, and Pearson Correlation for All Variables.

Variable	M	SD	1	2	3	4	5	6	7	8	9	10	11	12
1. Gender	0.61	0.49
2. Age	23.30	1.04	−.15**
3. ccn	3.61	0.65	.06	.03
4. ctr	3.91	0.60	.02	−.08*	.49**
5. cur	3.58	0.56	.02	−.02	.67**	.62**
6. cfd	3.68	0.58	.06	−.06	.66**	.69**	.73**
7. CA	3.69	0.51	.05	−.04	.83**	.82**	.88**	.90**
8. pv	4.86	1.40	−.11**	.00	−.12**	−.11**	−.08*	−.07	−.11**
9. anx	5.63	1.47	−.06	−.02	−.09*	−.10*	−.11**	−.09*	−.11**	.55**
10. si	5.09	1.64	−.09*	.03	−.14**	−.21**	−.15**	−.17**	−.19**	.59**	.59**
11. EPCD	5.20	1.28	−.10**	.01	−.14**	−.17**	−.14**	−.13**	−.17**	.83**	.84**	.87**
12. HS	5.10	0.84	.05	−.03	.47**	.31**	.41**	.42**	.47**	.00	.10*	.04	.05
13. D	4.20	1.05	−.15**	−.01	.11**	−.06	.06	−.02	.03	.34**	.31**	.44**	.43**	.43**
HS*	—	—	—	—	.47***	.38***	.42***	.46***	.50***	−.16***	−.04	−.18***	−.15***
D*	—	—	—	—	−.12**	−.23***	−.14***	−.24***	−.22***	.38***	.30***	.47***	.45***

Note. HS* = the partial correlations between high standard and other career indices when controlling discrepancy; D* = the partial correlations between discrepancy and other career indices when controlling high standard; ccn = concern; ctr = control; cur = curiosity; cfd = confidence; CA = career adaptability; pv = pessimistic views; anx = anxiety; si = self-concept and identity; EPCD = emotional and personality-related career difficulties; HS = high standard; D = discrepancy.

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

Grouping Participants by Cluster Analysis

The three-cluster solution and four-cluster solution generated from the traditional two-step cluster analysis procedure are presented in Table 2. As a result, there was no cluster within the three-cluster solution was identified as consisting of the high adaptive and low maladaptive dimensions representative for the HPs subtype of the tripartite model. However, the four-cluster solution supported the classification hypotheses of the 2 × 2 model. Based on the pairwise comparisons between the clusters, C1 to C4 was determined to correspond with within MP, NP, PHS, and PD within the 2 × 2 model, respectively.

Table 2.

The Three-Cluster and Four-Cluster Solution From Two-Step Cluster Analysis.

Dimension	Three-Cluster Solution			Four-Cluster Solution
	Cluster 1 (n = 278)	Cluster 2 (n = 115)	Cluster 3 (n = 279)	Cluster 1 (n = 240)	Cluster 2 (n = 88)	Cluster 3 (n = 136)	Cluster 4 (n = 208)
	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)
HS	0.71_a (.67)	−1.54_b (.50)	−0.07_c (.61)	0.89_a (.53)	−1.60_b (.54)	0.24_c (.63)	−0.51_d (.47)
D	0.89_a (.64)	−0.78_b (.77)	−0.57_c (.62)	0.90_a (.71)	−1.04_b (.65)	−1.08_b (.44)	0.11_c (.45)

Note. All F tests among clusters were significant at p < .001. Means that were significantly different (p < .05) in the post hoc analysis are indicated by different lettered subscripts. HS = standardized score of high standard; D = standardized score of discrepancy.

Grouping Participants by LPA

Comparing the models examined in the explorative and confirmatory LPA (see Table 3), the four-class model with fixed group means that derived from the two-step cluster analysis was selected as the optimal model. After excluded the model with inadequate class size (i.e., the five-class model in the explorative LPA), the four-class model with fixed group means had moderately low AIC, BIC, and a-BIC, significant p values in both LRT and BLRT tests, the highest entropy, and a meaningful classification structure. Overall, the 2 × 2 model was verified in LPA and then used for further analysis to investigate the relationship between perfectionism subtypes and career development.

Table 3.

Model Fit Results of the Explorative and Confirmatory LPA.

Model	df	AIC	BIC	a-BIC	LRT	BLRT	Entropy	Class Sizes
1. Class model	4	3677.75	3695.65	3682.95	-	-	-	648
2. Class model	7	3559.16	3590.48	3568.25	p < .0001	p < .0001	.57	246 402
3. Class model	10	3532.82	3577.56	3545.81	.04	p < .0001	.56	95 386 167
4. Class model	13	3520.49	3578.65	3537.37	.46	p < .0001	.60	69 311 194 74
5. Class model	16	3517.31	3588.89	3538.09	.17	.10	.60	15 168 97 200 168
4. Class model with fixed group means	11	3531.10	3580.32	3545.39	p < .0001	p < .0001	.65	214 65 122 247
3. Class model with fixed group means	8	3550.90	3586.69	3561.29	p < .0001	p < .0001	.42	262 169 217

Note. LPA = latent profile analysis; AIC = Akaike information criterion; BIC = Bayesian information criterion; a-BIC = adjusted Bayesian information criterion; LRT = Lo-Mendell-Rubin likelihood ratio test; BLRT = bootstrap likelihood ratio test. The boldface indicates the selected model.

Career Indices and Perfectionism Subtype Comparisons

As the result of the BCH method (see Table 4), NP had no significant difference from PD for career adaptability, χ²(1) = 3.01, p > .05, but was significantly lower than PD for career decision-making difficulties, χ²(1) = 14.82, p < .001, refuting Hypothesis B1a and validating Hypothesis B1b. MP was significantly higher than PD for career adaptability, χ²(1) = 54.11, p < .001, but had no significant difference for career decision-making difficulties, χ²(1) = 0.261, p > .05, validating hypothesis B2a but refuting Hypothesis B2b. PHS had no significant difference from MP for career adaptability, χ²(1) = 1.80, p > .05, but was significantly lower than MP for career decision-making difficulties, χ²(1) = 46.89, p < .001, refuting Hypothesis B3a and validating Hypothesis B3b. Exploration of the unconfirmed relationship between NP and PHS uncovered no significant difference for career decision-making difficulties, χ²(1) = 2.71, p > .05, but that PHS was higher than NP for career adaptability, χ²(1) = 39.34, p < .001. No two subtype clusters demonstrated completely similar or indistinguishable effects for career development outcome variables of career adaptability and career decision-making difficulties, indicating the potential use of the four clusters in the 2 × 2 model to distinguish individuals by theoretical career outcomes according to subtype.

Table 4.

The Relationships Between Perfectionism Subtypes and Career Indices.

Dimension	MP (n = 214)		NP (n = 65)		PHS (n = 122)		PD (n = 247)
Dimension	M	SD	M	SD	M	SD	M	SD
CA	0.71_a	0.51	.00_b	.43	.82_a	.38	0.19_b	.25
EPCD	1.27_a	1.10	.41_b	.86	.00_b	.97	1.20_a	.44

Note. Means that were significantly different (p < .001) in the Wald test are indicated by different lettered subscripts. CA = latent variable of career adaptability; EPCD = latent variable of emotional and personality-related career decision-making difficulties; MP = mixed perfectionism; PHS = pure high standard; PD = pure discrepancy.

Discussion

To determine whether the tripartite model or 2 × 2 model of perfectionism would be more relevant for career development of Chinese undergraduates, cluster analysis and LPA was applied to group participants which demonstrated the 2 × 2 model was more accurate for representing the population and outcomes in this study. Then, BCH approach was used to compare perfectionism subtypes with career development indices of positive career adaptability and negative career decision-making difficulties. The 2 × 2 model proved potent in identifying whether students possessed adaptive social–psychological resources conducive to career advancement or various risk factors exasperating potential for career problems according to their designated perfectionism subtypes.

Classification of Subtypes

Both traditional cluster analysis method and the relatively newly developed LPA method revealed the Chinese undergraduates in this study could all be aligned with four perfectionism subtypes indicative of the 2 × 2 model. These results were consistent with the results of a study for Chinese information technology employees (Li et al., 2014) which replicated the measurements of perfectionism and cluster analysis procedure used in this study. However, another study conducted in the United States (Rice & Ashby, 2007) also replicated the measurements of perfectionism and cluster analysis procedure used in this study but found the tripartite model as most relevant for university students. In other words, there is a low standards and high-discrepancy subtype in Chinese undergraduate and working adult population while not in American undergraduate population.

The inconsistency between these results may arise from the cultural disparities pertaining to standards of success and participants’ self-report tendency. On the one hand, social evaluation, interpersonal relationships, and peer expectations are emphasized in Chinese culture (Hwang, 2006) and thus individuals with Chinese heritage are more likely to set goals and standards oriented toward social desirability even when they conflict with individual abilities, interests, or values. This universal external-oriented standards of success may result in a sense of discrepancy even when personal standards are low. On the other hand, it is self-report tendency for Chinese to say “My self-standards are not high enough” because it is a performance of modesty since it hints that “I am not good enough” and modesty is a significant virtue in China. Thus, the low standards reported by Chinese participants may be underrated and not as reachable as it sounds to be. Therefore, individuals with Chinese heritage are more likely to identify with a PD subtype of perfectionism found only in the 2 × 2 model.

The Functions of Perfectionism Subtypes

This study supported the hypotheses of the 2 × 2 model that the PD subtype was most detrimental as it had lowest career adaptability and highest career decision-making difficulties, and the PHS subtype was most functional as it had the highest career adaptability and lowest career decision-making difficulties. The tripartite model was refuted because the MP subtype was not the most dysfunctional one since it had high career adaptability.

Furthermore, what need to be noticed is that the function of the subtypes is not inherent (Gaudreau, 2013); it depends on the criteria of function, namely, the outcome variables and the number of the criteria taken into consideration. Taking the result of this study, for example, for career adaptability, MP is as salutary as PHS, while for career decision-making difficulties, MP is more harmful; when consider only one criterion career adaptability, NP is as detrimental as the PD, while when consider both two career indices, PD is the most detrimental one. That suggested that the conclusion of the function of perfectionism subtypes cannot be analogized from one outcome domain to another one.

Practical Implications for Each Perfectionism Subtype

The PHS subtype with dimensions of high career adaptability and low decision-making difficulty is identified as the healthiest perfectionism subtype. Because of the capabilities already manifest, career counseling could prioritize the focus on skills training (i.e., mock interviews) and providing information on career development opportunities to actualize potential in lieu of focus on regulating emotion or building self-potency.

While individuals of the MP subtype possess the capacity to activate adaptive social–psychological resources, they also possess significant risk of career decision-making difficulties. Career intervention for individuals of this subtype could be optimized with locating their resources and helping them to realize their positive intrinsic tendencies needed to cope with career barriers. What’s more, the deep-seated internal conflict between elevated self-standards and inability to meet such standards may be critical source of their career problems. The counseling-based career guidance (Niles & Harris-Bowlsbey, 2013) could be an ideal strategy to alleviate such conflict.

The NP subtype demonstrates less risk from a low degree of career decision-making difficulties but also a low degree of career adaptability. These individuals may identify as somewhat apathetic or motiveless for self-improvement because lowered self-standards obscure their perception of any salient problems. Therefore, strategies for career counseling could include active outreach, facilitating awareness on the need for adaptability, and then helping them accumulate intrinsic resources as they become aware of their need.

The PD subtype with dimensions of low career adaptability and high career decision-making difficulty is identified as the unhealthiest subtype of perfectionism. Career intervention could be most impactful with early detection and preventative measures. One possibility is the use of prescreen procedures in schools to identify students at risk, so that more intensive and targeted career guidance can be provided before problems take root and become significantly detrimental.

Limitations and Future Directions

Two limitations were present in this study and should be considered when evaluating the findings. Firstly, the use of cross-sectional design with generalized measurement of perfectionism may reduce the ecological validity of the study. Although perfectionism is considered a stable personality trait (Stoeber & Otto, 2006), both stable and flexible components of perfectionism exist and could vary in their presentation depending on environmental context with change over time (i.e., repeated mistakes leading to stricter or looser standards). Thus, inclusion of cross-sectional design with generalized measurement of perfectionism could only recognize a moment in time for complex individuals that fluctuate in their presentation of perfectionism, rendering their determined categorizations for this study. Conducting studies with longitudinal and multilevel designs presents as a consideration for future research to tackle these concerns. Secondly, the reductionist nature of the quantitative analysis used in this study fundamentally restricts information that can be elucidated about the cognitive and behavioral processes involved in perfectionism. Future studies could consider the use of mixed methods to garner a more complete picture of perfectionism and associated subtypes.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Social Science Foundation of China (grant numbers 15ZDB139).

ORCID iDs

Danni Wang

Zhi-Jin Hou

Note

References

Andrews

L. M.

Bullock-Yowell

Dahlen

E. R.

Nicholson

B. C.

(2014). Can perfectionism affect career development? Exploring career thoughts and self-efficacy. Journal of Counseling and Development, 92, 270–279. doi:10.1002/j.1556-6676.2014.00155.x

Arana

F. G.

Furlan

(2016). Groups of perfectionists, test anxiety, and pre-exam coping in Argentine students. Personality and Individual Differences, 90, 169–173. doi:10.1016/j.paid.2015.11.001

Asparouhov

Muthén

B. O.

(2014). Auxiliary variables in mixture modeling: Using the BCH method in Mplus to estimate a distal outcome model and an arbitrary second model. Version 2. Retrieved from http://www.statmodel.com/examples/webnotes/webnote21.pdf

Blankstein

K. R.

Dunkley

D. M.

(2002). Evaluative concerns, self-critical, and personal standards perfectionism: A structural equation modeling strategy. In Flett

G. L.

Hewitt

P. L.

(Eds.), Perfectionism: Theory, research, and treatment (pp. 285–315). Washington, DC, US: American Psychological Association. doi:10.1037/10458-012

Bakk

Vermunt

J. K.

(2016). Robustness of stepwise latent class modeling with continuous distal outcomes. Structural Equation Modeling: A Multidisciplinary Journal, 23, 20–31. doi:10.1080/10705511.2014.955104

Cox

B. J.

Enns

M. W.

Clara

I. P.

(2002). The multidimensional structure of perfectionism in clinically distressed and college student samples. Psychological Assessment, 14, 365–373. doi:10.1037/1040-3590.14.3.365

Gati

Krausz

Osipow

S. H.

(1996). A taxonomy of difficulties in career decision making. Journal of Counseling Psychology, 43, 510–526. doi:10.1037/0022-0167.43.4.510

Gaudreau

(2013). The 2 × 2 model of perfectionism: Commenting the critical comments and suggestions of Stoeber (2012). Personality and Individual Differences, 55, 351–355. doi:10.1016/j.paid.2013.03.021

Gaudreau

Thompson

(2010). Testing a 2 × 2 model of dispositional perfectionism. Personality and Individual Differences, 48, 532–537. doi:10.1016/j.paid.2009.11.031

10.

Gaudreau

Verner-Filion

(2012). Dispositional perfectionism and well-being: A test of the 2 × 2 model of perfectionism in the sport domain. Sport, Exercise, and Performance Psychology, 1, 29–43. doi:10.1037/a0025747

11.

Grzegorek

J. L.

Slaney

R. B.

Franze

Rice

K. G.

(2004). Self-criticism, dependency, self-esteem, and grade point average satisfaction among clusters of perfectionists and nonperfectionists. Journal of Counseling Psychology, 51, 192–200. doi:10.1037/0022-0167.51.2.192

12.

Hair

L. F.

Black

W. C.

Babin

B. J.

Anderson

R. E.

(2009). Multivariate data analysis (7th ed.). Cranbury, NJ: Pearson Education.

13.

Hamachek

D. E.

(1978). Psychodynamics of normal and neurotic perfectionism. Psychology: A Journal of Human Behavior, 15, 27–33. Retrieved from https://search.proquest.com/docview/616292295?accountid=8554

14.

Hewitt

P. L.

Flett

G. L.

(1991). Perfectionism in the self and social contexts: Conceptualization, assessment, and association with psychopathology. Journal of Personality and Social Psychology, 60, 456–470. doi:10.1037/0022-3514.60.3.456

15.

Hou

Leung

S. A.

(2012). Career adapt-abilities scale—China form: Construction and initial validation. Journal of Vocational Behavior, 80, 686. doi:10.1016/j.jvb.2012.01.006

16.

Hwang

(2006). Moral face and social face: Contingent self-esteem in Confucian society. International Journal of Psychology, 41, 276–281. doi:10.1080/00207590544000040

17.

Ikeda

Hou

Liu

Itamar

(2016). Validation of the emotional and personality-related career decision-making difficulties questionnaire short form (EPCD-SF) in Chinese university students. Chinese Journal of Clinical Psychology, 24, 444–449.

18.

Lehmann

I. S.

Konstam

(2011). Growing up perfect: Perfectionism, problematic internet use, and career indecision in emerging adults. Journal of Counseling and Development, 89, 155–162.

19.

Lent

R. W.

(2005).A social cognitive view of career development and counseling. In Brown

S. D.

Lent

R. W.

(Eds.), Career development and counseling: Putting theory and research to work (pp. 101–127). Hoboken, NJ: John Wiley.

20.

Leong

F. T. L.

Chervinko

(1996). Construct validity of career indecision: Negative personality traits as predictors of career indecision. Journal of Career Assessment, 4, 315–329. doi:10.1177/106907279600400306

21.

Hou

Z.-J.

Chi

H.-Y.

Liu

Hager

M. J.

(2014). The mediating role of coping in the relationship between subtypes of perfectionism and job burnout: A test of the 2×2 model of perfectionism with employees in China. Personality and Individual Differences, 58, 65–70. doi:10.1016/j.paid.2013.10.007

22.

Mobley

Slaney

R. B.

Rice

K. G.

(2005). Cultural validity of the almost perfect scale-revised for African American college students. Journal of Counseling Psychology, 52, 629–639. doi:10.1037/0022-0167.52.4.629

23.

Muthén

L. K.

Muthén

B. O.

(2015). MPlus user’s guide (7th ed.). Los Angeles, CA: Author.

24.

Niles

S. G.

Harris-Bowlsbey

(2013). Introduction to career development interventions. In Niles

S. G.

Harris-Bowlsbey

(Eds.), Career development interventions in the 21th century (4th ed., p. 27). New York, NY: Pearson Education.

25.

Nylund

K. L.

Asparouhov

Muthén

B. O.

(2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling, 14, 535–569.

26.

Parker

W. D.

Adkins

K. K.

(1995). A psychometric examination of the multidimensional perfectionism scale. Journal of Psychopathology and Behavioral Assessment, 17, 323–334. doi:10.1007/BF02229054

27.

Page

Bruch

M. A.

Haase

R. F.

(2008). Role of perfectionism and five-factor model traits in career indecision. Personality and Individual Differences, 45, 811–815. doi:10.1016/j.paid.2008.08.013

28.

Periasamy

Ashby

J. S.

(2002). Multidimensional perfectionism and locus of control: Adaptive vs. maladaptive perfectionism. Journal of College Student Therapy, 17, 75–86. doi:10.1300/J035v17n02_06

29.

Rice

K. G.

Ashby

J. S.

(2007). An efficient method for classifying perfectionists. Journal of Counseling Psychology, 54, 72–85. doi:10.1037/0022-0167.54.1.72

30.

Rice

K. G.

Ashby

J. S.

Slaney

R. B.

(1998). Self-esteem as a mediator between perfectionism and depression: A structural equations analysis. Journal of Counseling Psychology, 45, 304–314. doi: 10.1037/0022-0167.45.3.304

31.

Rice

K. G.

Ashby

J. S.

Slaney

R. B.

(2007). Perfectionism and the five-factor model of personality. Assessment, 14, 385–398. doi:10.1177/1073191107303217

32.

Rheaume

Freeston

M. H.

Dugas

M. J.

Letarte

Ladouceur

(1995). Perfectionism, responsibility, and obsessive–compulsive symptoms. Behaviour Research and Therapy, 33, 785–794. doi: 10.1016/0005-7967(95)00017

33.

Saka

Gati

Kelly

K. R.

(2008). Emotional and personality-related aspects of career-decision-making difficulties. Journal of Career Assessment, 16, 403–424. doi:10.1177/1069072708318900

34.

Savickas

M. L.

(1997). Career adaptability an integrative construct for life-span, life-space theory. The Career Development Quarterly, 45, 13.

35.

Savickas

M. L.

(2013). Career construction theory and practice. In Brown

S. D.

Lent

R. W.

(Eds.), Career development and counseling: Putting theory and research to work (2nd ed., pp. 157–161). Hoboken, NJ: John Wiley.

36.

Savickas

M. L.

Porfeli

E. J.

(2012). Career adapt-abilities scale: Construction, reliability, and measurement equivalence across 13 countries. Journal of Vocational Behavior, 80, 661–673. doi:10.1016/j.jvb.2012.01.011

37.

Slaney

R. B.

Rice

K. G.

Mobley

Trippi

Ashby

J. S.

(2001). The revised almost perfect scale. Measurement and Evaluation in Counseling and Development, 34, 130–145.

38.

Smith

M. M.

Saklofske

D. H.

Yan

Sherry

S. B.

(2015). Perfectionistic strivings and perfectionistic concerns interact to predict negative emotionality: Support for the tripartite model of perfectionism in Canadian and Chinese university students. Personality and Individual Differences, 81, 141–147. doi:10.1016/j.paid.2014.09.006

39.

Stoeber

(2012). The 2 × 2 model of perfectionism: A critical comment and some suggestions. Personality and Individual Differences, 53, 541–545. doi:10.1016/j.paid.2012.04.029

40.

Stoeber

Mutinelli

Corr

P. J.

(2016). Perfectionism in students and positive career planning attitudes. Personality and Individual Differences, 97, 256–259. doi:10.1016/j.paid.2016.03.065

41.

Stoeber

Otto

(2006). Positive conceptions of perfectionism: Approaches, evidence, challenges. Personality and Social Psychology Review, 10, 295–319. doi:10.1207/s15327957pspr1004_2

42.

Yang

Liang

B. Y.

Zhang

X. G.

Y. C.

(2007). The Chinese version of the almost perfect scale-revised. Studies of Psychology and Behavior, 5, 139–144.