UNIVERSITAS INDONESIA ALAT UKUR BARU PERFEKSIONISME MAKALAH NON SEMINAR Stevie Kurnia 0906552340 Fakultas: Psikologi Program: Kelas Khusus Internasional Depok Januari, 2014
Alat Ukur Baru Perfeksionisme 1,2 Stevie Kurnia, 1 Mark Horswill, 1 Riyad Rahimullah, 2 Donny Hendrawan 1. Undergraduate Programme, Faculty of Arts, University of Queensland, St. Lucia, Brisbane, 4078, Australia 2. Program Sarjana, Fakultas Psikologi, Universitas Indonesia, 16424, Depok, Indonesia E-mail: kurniastevie@gmail.com Abstrak Riset ini bertujuan untuk menciptakan skala perfeksionisme yang baru dengan reliabilitas dan validitas yg baik. Skala baru tersebut mempunyai 10 item dengan menggunakan skala likert 5 poin. Partisipan dari riset ini adalah mahasiswa University of Queensland yang terdaftar di dalam mata kuliah PSYC 3020 ( N =246) dengan 176 partisipan wanita dan 70 partisipan pria dengan rata rata umur ( M =22.071, SD =4.307) Internal konsistensi skala baru ini diuji dengan Cronbach s alpha yang menunjukan internal konsistensi ( α =. 6 0 8). Untuk validitas, skala baru ini mempunyai hubungan positif dengan s kala Almost Perfect Scale -Revised ( APS -R ) oleh Slaney, Rice, Mobbley, Trippi, & Ashby (2001), namun berkolerasi negatif dengan skala self- esteem Rosenberg (1965). Tidak ditemukan korelasi yang signifikan antara skala baru dengan I nternational P ersonality I tem P ool (IPIP) conscientiousness scale. Skala baru ini dapat diaplikasikan ke dalam kehidupan kerja untuk keperluan rekrutmen di beberapa bidang pekerjaan. The new perfectionism scale Abstract The study aims to create a new scale of perfectionism with a good reliability and validity. The scale has 10 items with 5 point likert scale. The participants of this study were the University of Queensland students who are enrolled in PSYC3020 course ( N =246) with 176 female participants and 70 male participants with average age of ( M =22.071, SD =4.307) Internal consistency of the scale was tested and it revealed that the cronbach s alpha of the scale was ( α =. 608 ). In terms of validity, the scale had a positive correlation with ASP-R scale by Slaney, Rice, Mobbley, Trippi, & Ashby (2001) and a negative correlation with Rosenberg s (1965) self-esteem scale. No significant correlation found between the new scale and the International Personality Item Pool (IPIP) conscientiousness scale. The scale could be used in daily life as a recruitment tool in several job fields Keywords: Perfectionism; conscientiousness; self-esteem new scale; APS-R scale Introduction Mitchelson (2009) described perfectionism as a trait consists of three dimensions which are high personal standards, the need for order, and discrepancy between stand ards and perceived performance. Other study by Slaney, Rice, and Ashby (2002) clustered perfectionism to two things. The first thing being adaptive perfectionism (low discrepancy between personal high standard and actual performance) and maladaptive perfectionism (high discrepancy between personal high standard and perceived performance). Perfectionism is a
common trait that we could easily find in daily life. Moreover, perfectionism could lead or predict to other traits or behaviour. Several studies have revealed relationships between perfectionism and few factors in life such as work family conflict, self-esteem, and several psychological problems. Mitchelson (2009) in her study found to have higher discrepancy between standards and behaviour in home domain. Women were also found to have higher standards. The other findings that Mitchelson (2009) found is that ambiguous expectations at home work against perfectionist. In the study, participants also reported to have higher discrepancy and higher standards at home. The study also found that adaptive perfectionists have lower perceived work family conflict than maladaptive perfectionists and non-perfectionists. The study also suggested that self-efficacy has a role in reducing work family conflict. A nother study by Wang, Slaney, and Rice (2007) discussed about perfectionism and self-esteem among Chinese university students. The study also used three groups to be compared; adaptive perfectionists, maladaptive perfectionists, and non-perfectionists. Among those groups, participants in the maladaptive perfectionists scored lower on self-esteem and higher on anxiety and depression. Adaptive perfectionists group scored higher on self-esteem and lower on anxiety and depression while non-perfectionists scored between those two groups. This study uniquely found a fourth group where they scored lower on state anxiety but had similar scores for trait anxiety and depression to maladaptive perfectionists. However, those scores were higher than the adaptive perfectionists. Those participants on the fourth group scored lower on self-esteem than the adaptive perfectionists and in between of maladaptive perfectionists and non-perfectionists. The existence of the fourth group in this study is unique as the number of participants who belong in this group was high. In terms of motivation, the fourth group scored lower than the perfectionists groups but similar to the non -perfectionists. However, in terms of social achievement motivation, the fourth group scored was similar to the perfectionists groups (higher than the non-perfectionist group). In relation between perfectionism and psychological problems, Rice and Pence Jr. ( 2006 ) investigated the relationship between perfectionism and Obsessive Compulsive Disorder ( OCD ) symptoms. The study examine d whether A lmost P erfect S cale -R evised (APS- R) constructed by Slaney, Rice, Mobbley, Trippi, & Ashby ( 2001 ) could be used as a predictor of OCD symptoms. The study found that the only thing in the ASP-R scale that accurately predicted OCD symptoms was discrepancy. The second finding of the study is that participants with elevated personal standards was buffered from the increased symptoms of OCD compared to participants with lower personal standards.
The present study examined a new measurement tool for personality trait of perfectionism. Based on the studies reviewed above, the most widely used scale to measure perfectionism is the Almost Perfect Scale (APS) by Slaney and Johnson (1992) and the APS- R by S laney et al. (2001). The scale was found to have a good reliability, by having high Cron bach s alpha scores ranging from.82 to.92. In order to examine the validity, the study used construct validity by looking at the correlations of selected subscales within each samples. The correlations was found to be in expected directions and they were found to be moderate to high correlations. This scale will improve t he studies of perfectionism by providing an option of scale for measuring perfectionism. We have developed a new scale for perfectionism by having a pilot study with small samples. The items in the scale were including some items from Slaney et al. (2001), APS-R, with rewording or removal some of the items in the APS-R. The research design was the sa me with the ASP-R scale which was a correlational design. The data from the participants were taken once online. The new perfectionis m scale is worth measuring since there is not many perfectionism scale s from previous studies. The new perfectionism scale could provide new options for measuring perfectionism in future studies. In order to test the reliability of this new scale, we examined the inter nal consistency by measuring the Cronbach s alpha coefficient. This scale will be considered reliable if it has a high score of Cronbach s alpha. The validity of this test will be tested by generating three hypotheses. The first hypotheses is that if the new perfectionism is valid, there will be a high correlation found between the new scale of perfectionism and t he ASP-R scale by Slaney et al. (2001) as the ASP-R scale also measures perfectionism. The second hypothesis is that this scal e would have a high correlation to the Goldber g, Johnson, Eber, Hogan, Ashton, Cloninger, & Gough (2006) IPIP (International Personality Item Pool) c onscientiousness scale. The IPIP conscientiousness scale would be a good validating scale to the new perfectionism scale as previous study by Rice, Ashby, and Slaney (2007) about five factor model and perfectionism showed that conscientiousness had a high association with perfectionism. Another study by Stoeber, Otto, & Dalbert (2009) found that conscientiousness predicted a longitudinal increases for self-oriented perfectionism among adolescents. The third hypothesis is that the new scale would be highly correlated with a self-esteem inventory scale by Rosenberg (1965). Previous study by Slaney et al., (2001) about the ASP-R found negative correlation between discrepancy and GPA and self-esteem.
Method Participants and procedures The par ticipants of this study were 246 students ( M =70, F =176 ) who were undergraduate students who currently enrolled in PSYC3020 with age range from 19-56 years old ( M =22.071, SD =4.307 ). Participants were recruited during the PSYC3020 tutorial class and before commencing the data gathering, participants were told that their involvement was voluntary. Along with the questionnaire, participants demographic such as age and gender were also asked. Participants filled the online questionnaire with the help of their tutors. Participants data were collected online from all tutorial classes of PSYC3020 course. The data collected online were stored in an online database to be processed. Measures The new pe rfectionism scale consists of 10 items. The new perfectionism scale will figure out how perfectionists the participants are. The wording and selections of items were based on the agreement from the researchers which took place in PSYC 3020 tutorial classes. The items in the scale include When I complete a task, I feel I could do more 2 out of 10 items were reversed in order to maintain the consisten cy of participan ts responses. Such item is I am satisfied with my current level of achievement. The rating of the items use 5 points likert scale with 1 (strongly disagree) to 5 (strongly agree) with the scale midpoint of 3 (neutral). The internal consistency of this scale was found to be not strong enough with ( α =.608) (See appendix). Three validating scales were used in order to check the validity of the new perfectionism scale (See appendix). The first validating scale is APS -R scale by Slaney et al., (2001) which covers perfectionism. The APS-R scale has no items to be scored reversely. The questions in the scale include I have high standards for my performance at work or at school. The APS -R scale has a good reliability with Cronbach s alpha ranged from.82 to.92. The second validating scale is the conscientiousness part of the IPIP scale by Goldberg et al., (2006). The conscientiousness part of the scale include questions such as Am always prepared. There are several items on the scale that are reversed such as Find it difficult to get down to work.
The third validating scale is the self-esteem scale by Rosenberg (1965). The scale include such questions such as All in all, I am inclined to feel that I am a failure. The scale also include several reversed scoring items such as I feel I do not have much to be proud of. Results Descriptive statistics revealed the mean, standard deviation, skewness, and range (See table 1). The data for such statistics were normally distributed. Table 1. Descriptive statistics for new perfectionism, APS-R, NEO-PI-R (conscientiousness), and Rosenberg s self-esteem scales M SD Minimum Maximum Skew New perfectionism 31.96 4.672 20 47.504 ASP-R 41.22 7.19 24 60.353 conscientiousness self esteem 32.63 24.91 6.48 1.82 10 19 49 30 -.205 -.103 Pearson s correlation was used in order to test the correlation of the new perfectionism scale with the validating scales as the data was normally distributed and the variables were continuous. T here was a positive correlation (r =.607, p <.001) b etween the new perfectionism scale and the ASP-R scale ). The correlation showed that the higher participants scored in the new perfectionism scale, they would be more likely to score high in the ASP-R scale. However, there was no significant correlation between the new perfectionism scale and conscientiousness scale ( r =.37, p =.58). In terms of the correlation between the new perfectionism scale and the self-esteem scale, a significant negative correlation was found ( r = -.185, p =.005). Such correlation revealed that if participants scored highly on the new perfectionism scale, they would get a low score in the self-esteem scale. Table 2. Item Discrimination Indices for the new perfectionism scale L U d Item 1 21 48.40 Item 2 98 101.26 Item 3 18 50.46 Item 4 15 41.38 Item 5 15 41.38 Item 6 33 61.44 Item 7 27 63.53 Item 8 10 38.39 Item 9 11 44.45 Item 10 13 50.51
Note. n u = 72 and n L= 96 In order to test the discrimination between the high scores and the low scores on the new perfectionism scale, item discrimination indices ( d ) was used. Item discrimination indices were calculated by subtracting the numbers of high performers ( U ). The numbers then divided by the high scores ( nu ) and the numbers of low performers who got some questions right ( L) divided by total number of low performers ( nl ). The item discrimination indices showed that participants scored the highest on item 7 ( d =.53). On the other hand, participants scored the lowest on item 2 (d =.26) Discussion The aim of this study is to test the reliability and the validity of the new perfectionism scale. From the statistics above, there are several things that the scale found abo ut perfectionism related to the hypotheses made in the study earlier. The first findings of the study was that the new perfectionism scale was po sitively correlated with the APS -R by Slaney et al., (2001). The first finding supported the hypothesis of the study. The positive correlation between the two scales indicated that participants would get high score s on the new perfectionism scale if they got high scores on the APS -R scale. The possible reason of the positive correlation between the two scales was because both new perfectionism scale and the APS-R scale measured the same thing, which is perfectionism. The second finding of the study revealed that there was no significant correlation found between the new perfectionism scale and the IPIP conscientiousness scale. Unfortunately, the second finding did not support the second hypothesis where it was suggested that the new perfectionism scale would have a positive correlation with the IPIP conscientiousness scale. The finding of this study on perfectionism and conscientiousness was different from previous study by Stoeber et al. (2009) where they found a positive correlation between conscientiousness and self-oriented perfectionism. The difference in findings between both studies might be caused by several factors. The first factor might be the design difference between both studies. The study by Stoeber et al. (2009) was a longitudinal study where this study was a correlational study. The second factor might be age difference of the participants in both studies. Stoeber et al. (2009) s participants were younger people who aged from 14-19 years old with 5-8 months period. The average age of this study s participants were 22 years old and the test was taken in one sitting. This factor might affect
the results of the study in a way that fatigue might affect participants performances and small range of age might also affect participants performances. The third finding of this study was there was a significant negative correlation between the new perfectionism scale and the Rosenberg s (1965) self-esteem scale. The negative correlation found between both scales indicated that participants who got high scores on the new perfectionism scale would get low scores on the self-esteem scale. The third finding w as similar to previous study by Slaney et al., (2001) where they found that there was a negative correlation between self-esteem and discrepancy, which was a part of perfectionism among college students. Similar findings also found in a study by Gotwals & Dunn (2003) to student athletes. The possible explanation is that perfectionism happened because of the lack of self-esteem. One could be a perfectionist in order to increase his selfesteem. The reliability of the new perfectionism scale was found to be weak with cronbach s alpha of.608. Unfortunately, the reliability of this new perfectionism scale was weaker than the ASP-R scale which had cronbach s alpha ranged from.82 to.92. In terms of items discrimination, participants scored highest on item 7 I strive for high standards even if it costs me to achieve it The item number 7 had the participants to score highly might be caused by the wording in the item which directly describe a perfectionist trait. On the other hand, participants scored the lowest on item number 2 When I complete a task, I feel I could do more The low scores by participants on item number 2 might be caused by ambiguity of the direction of the question so that participants were not sur e what the statement meant. In order to increase item discrimination indices on item n umber 2, the item could be reworded with I feel I did not give my best after completing a task The strength of this new perfectionism scale is that the scale was found to be valid. Such validity was acquired by having most of this study s hypotheses supported. The other strength of this scale is that it was direct and practical by not having too many items. Having fewer items made it easier to gather data from participants as most participants seem reluctant to be volunteering in a study when faced with many items in a scale. The strength of this study was a practical, easy to do data collection by doing it online. Online data collection made the data collected from the participants easy to process. There are several limitation that this study faced in progress. The first limitation is that the new perfectionism scale was found to have a weak reliability with cronbach s alpha of.608. The cronbach s alpha score showed that the scale had to be improved for the scale to be used in daily life. The second limitation is that the study used a self-reported scale. Self-
reported scale is vulnerable to a phenomenon called social desirability bias. Social desirability is a phenomenon where participants respond to items in a scale in order to make them look good. The third limitation is that the scale was not found to be valid in terms of its corr elation with conscientiousness as the findings did not find a significant correlation between conscientiousness and perfectionism. The last limitation is that the participants only gathered from the university settings as all participants for this study were university students who enrolled in PSYC3020 course. With the participants recruited in one setting, the findings might not be accurate in other community outside university. Having university students as participants also limit the average age of the participants. The average age for this study was 22 years old, which makes the findings might not be accurate for older people in their 30s or 40s. Future study could improve the new perfectionism scale by improving the reliability of the new scale. Improvement in reliability can be done by addi ng more items to the scale to improve its internal consistency, and by doing a test retest reliability. Other than improve the scale s reliability, test retest reliability could help detecting any social desirability bias by participants. The validity of this study could be improved by comparing the scale to other scales with similar objectives. In order to improve this scale s insignificancy with conscientiousness, improvements in wording the items might help to increase the correlation between two scales. Improvement in wording the items could be done by matching the items and the objective of the scale. Those process might will improve in terms of its construct validity. Nonetheless, the scale could be used in daily life as a recruiting measurement tools in several job fields. The scale might be useful in police departments, fire fighter departments, or army recruitments to recruit their candidates. The scale would be a useful tool to prevent unfit candidates getting those jobs as police, fire fighter, and army jobs are jobs that are hard to do perfectly. Preventing perfectionist candidates to get the jobs might be a good idea to prevent them becoming stressed not doing a perfect job. On the other hand, the scale could be used by corporates to recruit perfectionist candidates in order to minimize turnovers by their employees. In conclusion, the study aimed to create a new scale of perfectionism with a good reliability and validity. Unfortunately, the scale has not reached a good reliability yet with cro nbach s alpha of.608. The scale had a positive correlation with other scale of perfectionism and a negative correlation with a self-esteem scale. However, no significant correlation between this scale and a conscientiousness scale. Overall, the scale has a reasonable validity
but it could be improved by increasing its reliability. The reliability of this scale could be done by adding more items, or doing a test retest reliability while increasing its reliability, it also prevents the social desirability bias. The validity of the scale could be improved by correlate the scale with other more conscientiousness related scales. It could also be improved by rewording the items in order to reach a good construct validity. References Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C., Cloninger, C. R., & Gough, H. G. (2006). The international personality item pool and the future of publicdomain personality measures. Journal of Research in Personality, 84-96. Gotwals, J. K., Dunn, J. G., & Waymenn, H. Y. (2003). An examination of perfectionism and self esteem in intercollegiate athletes. Journal of sport behaviour, 17. Mitchelson, J. K. (2009). Seeking the perfect balance: Perfectionism and work family conflict. Journal of Occupational and Organizational Psychology, 349-367. Rice, K. G., & Pence,Jr., S. L. (2006). Perfectionism and Obsessive Compulsive Disorder. Journal of Psychopathology and Behavioral Assessment, 103-111. Rice, K. G., Ashby, J. S., & Slaney, R. B. (2007). Perfectionism and the Five-Factor Model of Personality. Assessment, 385-398. Rosenberg, M. (1965). Society and adolscent self image. Princeton University Press. Slaney, R. B., & Johnson, D. G. (1992). The Almost Perfect Scale. Unpublished Manuscript. Slaney, R. B., Rice, K. G., Mobbley, M., Trippi, J., & Ashby, J. (2001). The Revised Almost Perfect Scale. Measurement and Evaluation in Counseling and Development, 130. Stoeber, J., Otto, K., & Dalbert, C. (2009). Perfectionism and the Big Five: Conscientiousness predicts longitudinal increases self oriented perfectionism. Personality and Individual Differences, 363-368. Wang, K. T., Slaney, R. B., & Rice, K. G. (2007). Perfectionism in Chinese university students from Taiwan: A study of psychological well-being and achievement motivation. Personal and Individual Difference, 1279-1290.
Appendix The new perfectionism scale items I am satisfied with my current level of achievement When I complete a task, I often feel I could do more I experience anxiety when my diet and exercise is disrupted APS-R scale I often feel frustrated because I can t meet my goals My best just never seems to be good enough for me I rarely live up to my high standards Doing my best never seems to be enough I am never satisfied with my accomplishments I like to always be organised and disciplined IPIP conscientiousness Am always prepared Make plans and stick to them Shirk my duties Rosenberg s self-esteem scale I feel that I am a person of worth, at least on an equal plane with I certainly feel useless at times At times, I feel I am no good at all.