Blog

Student ID:Psychometrics: 11Development and validation of the Maladaptive Perfectionism InventoryPerfectionism is a multidimensional construct conceptualised by high personal standards,concern over mistakes and a high level of self-critical evaluation (Frost, Marten, Lahart, &Rosenblate, 1990; Hill & Curran, 2015). It is generally acknowledged that perfectionismconsists of two over—arching dimensions that are broadly connotative of positive or negativeoutcomes, although they have been described and labelled in a number of slightly differentways (Stoeber & Otto, 2006).The adaptive dimension of perfectionism is concerned with high personal standards and selforiented (intrinsically motivated) perfectionism, whilst the maladaptive dimension is focusedon self-doubt about actions, concern over making mistakes, socially prescribed (extrinsicallymotivated) perfectionism and a perceived discrepancy between high expectations and actualachievements (Gotwals, Stoeber, Dunn, & Stoll, 2012; Stoeber & Otto, 2006). Thismaladaptive dimension has been found to be associated with a plethora of negative outcomessuch as psychological distress (James, Verplanken, & Rimes, 2015), depressive symptoms(Békés et al., 2015), eating disturbances (Muyan, Chang, Jilani, & Yu, 2015; Shanmugam &Davies, 2015) and avoidant coping strategies (Moroz & Dunkley, 2015).This paper reports on the development of the Maladaptive Perfectionism Inventory (MPI), anew measure of self-reported maladaptive perfectionism designed for adult use within thegeneral population.The maladaptive dimension of perfectionism has been associated with a number ofnegative outcomes (Hill & Curran, 2015). This paper reports. Limitationsregarding reliability testing and factor extraction were highlighted, andrecommendations made for future research.Keywords: perfectionism, psychometrics, self-reportStudent ID:Psychometrics: 12MethodDesignThe study employed a cross-sectional design, using online questionnaires.ParticipantsParticipants wereParticipantswere recruited via .MaterialsMPI item poolExisting literature on perfectionism was used to develop a conceptual framework containingfour domains of maladaptive perfectionismIn addition, approximately 25% of the items werenegatively-keyedTable 1Maladaptive perfectionism domains and facetsDomain FacetExternal orientation Need for validationMistrust of external validationIdentity relianceFear of negative evaluationPerceived high standards of othersIntrinsic worth Inability to recognise achievementsReliance on achievements for self-worthSelf-worth comparative to achievements of othersSelf-evaluation Impossibly high standardsSelf-critical evaluationConcern over mistakesImpact on life Inability to leave tasks unfinishedInability to switch off from workWork always first priorityStudent ID:Psychometrics: 13As such, the decision to use fiveoptions was felt to strike an appropriate balance.The response options were considered to represent a broad range of thedimension being measured and, as far as possible, to present psychologically equal differencesbetween each option. A mid-point option (‘Neither agree nor disagree’) was included, as theseThe item pool was then pre-tested on 10 participants using a variety of cognitive interviewingtechniques (Collins, 2003) to determine whether there were any problems with language oritem formatting. Revisions were made based on the feedback from pre-testing, resulting in apool of 24 items..Table 2Scales used to establish validity of the Maladaptive Perfectionism InventoryMeasure Reference Number ofitemsReliability(?)†Validity typeAPS-Ra Slaney, Rice, Mobley, Trippi,& Ashby (2001)23 .85 – .92 Concurrent criterionOCI-Rb Foa et al. (2002) 18 .81 – .93 ConvergentPSSc Cohen, Kamarck &Mermelstein (1983)10 .84 – .86 ConvergentSDRS-5d Hays, Hayashi & Stewart(1989)5 .66 – .68 Discriminanta Revised Almost Perfect Scaleb Revised Obsessive-Compulsive Inventoryc Perceived Stress Scaled Socially Desirable Response Set: 5-item survey† Reliability estimates taken from original paper, as citedAnalysisThe first phase of the analysis focused on item reduction.Student ID:Psychometrics: 14The items retained were then used to explore the dimensionality of the MPI by running anexploratory factor analysis using an orthogonal, Varimax, rotation.the reduced item measure.Construct validity was explored

by examining aIt was also expected that MPI scores would beunrelated to social desirability. Criterion validity was then assessed by correlating the MPIwith the APS-R, an existing measure of perfectionism that is widely used and extensivelyvalidated (Stoeber & Madigan, 2016). It was expected that the MPI would correlate stronglywith this measure.All tests used in this study produce scores that are, strictly speaking, at an ordinal level ofmeasurement.However, despite this, it is generally accepted thatLikert-type scales can be treated as interval levels of measurement, particularly when usingaggregated scores from multi-item scales (Furr, 2011) as is the case in the present study.Student ID:Psychometrics: 15ResultsAll participants completed the 24-item pool. Only two items (Q9 and Q12) had missing data,and this was from only one participant on each occasion.Item analysisItem analysis of the MPI item pool highlighted problems withThe internal consistency of this reduced 17-item poolItem-total statistics were examined to identify itemsThis led to the removal of a further 2 items, resulting in a reduced pool ofDimensionalityPAF was then used to explore the scaling properties of the 15-item measure.The scree plot (see Figure 1) also pointed towardsFig. 1. Scree plot of factor eigenvaluesStudent ID:Psychometrics: 16Table 3Criteria for reduction of MPI item poolNo. ItemValidnFacilityindex RangeItem-total rn = 1371† I get frustrated if I do not finish a task I have started 139 3.7 1-5a 0.372 Satisfactory work is not good enough 139 3.6 1-5 0.523 I feel I can switch off from my work easily, even if Ihaven’t finished a task139 3.45 1-5 0.374 I am a failure if I get negative feedback 139 3.18 1-5 0.665†b I often compare my own achievements to those ofothers139 3.69 1-5 0.536 People will view me negatively if my work is not perfect 139 3.21 1-5 0.697† I always feel my work can be improved 139 4.2 2-5a 0.358† I must put 100% effort into the tasks that I do 139 3.7 1-5a 0.439 My need for perfection has a negative impact on mylife138 2.69 1-5 0.6210 My work is never good enough, no matter how mucheffort I make139 2.63 1-5 0.6311 I feel pressure to be better than others in the tasks thatI do139 3.19 1-5 0.6112† Failing to complete a task rarely bothers me 138 3.53 2-5 0.4413 I feel that others expect me to be perfect 139 3.06 1-5 0.6414 Work is not always my first priority 139 2.53 1-5 0.4115† I try to avoid mistakes at all costs 139 3.77 2-5a 0.4616† I am not happy to produce “satisfactory” work 139 3.54 1-5a 0.5017 I expect to excel at everything I do 139 3.09 1-5 0.4118† Making mistakes is a sign of stupidity 139 2.21 1-5a 0.5419†b Being perfect means I will be accepted by others 139 2.39 1-5 0.6720 The mistakes I make are worse than those made byothers139 2.5 1-5 0.6021 I am only proud of my work if I get praise 139 2.64 1-5 0.5722 My personal reputation is not decided by how hard Iwork139 2.75 1-5 0.4723 People expect more of me because of my highstandards139 3.21 1-5 0.5924 It’s OK if I make a mistake 139 2.45 1-5 0.63a Two or more adjacent scale points showing 0.3a Self-Critical Evaluation subscaleb Impossibly High Standards subscaleTotal scale scores were then computed for the full MPI and the subscales to enable validitytesting of the measure. Descriptive statistics for these are illustrated in Table 5.Table 5Descriptive statistics for MPI and its subscalesPossible range Actual range??? SD Lower Upper Lower UpperTotal MPI 32.03 7.72 11 55 13 50Total SCE subscalea 13.39 4.71 5 25 5 24Total HIS subscaleb 18.64 4.29 6 30 8 28a Self-Critical Evaluationb Impossibly High Standards subscaleStudent ID:Psychometrics: 18ValidityPearson’s correlations between the MPI, the SCE subscale, the IHS subscale and the otherscales used to establish validity are detailed in Table 6.Construct validityDiscriminant validity was supported.Convergent validity was also supported,Concurrent criterion validityTotal MPI scores were found to be strongly related toContent validityThe content of the MPI is based on aTable 6Zero-order correlations between MPI, its subscales, and scales used for validation testingMeasures Total MPI SCE subscale IHS subscale4. APS-Ra .74*** .67*** .59***5. OCI-Rb .35*** .26** .34***6. PSSc .52*** .53*** .35***7. SDRS-5d -.01† .03† -.04†8. Age -.08† -.05† –.09†*** Significant at p