Minimum Average Partial Research Articles

How to measure the general Environmental Health Literacy

Abstract Background Many environmental factors, both natural or man-made, are associated with health risks. To make informed decisions about these potential risks, certain skills and competencies are necessary. The concept of Environmental Health Literacy (EHL) which integrates aspects of Science Literacy, Health Literacy and Environmental Literacy encompasse these competencies. To assess to what extent individuals or groups are environmental health literate, a general Environmental Health Literacy Scale is necessary. To the best of our knowledge, there is no validated instrument to measure general EHL. Methods Two phases of scale construction were conducted: 1) Item development with deductive methods which includes a literature review and assessment of existing scales resulting in two suitable studies for scale construction. Included items were translated to German language. 2) Scale development with pre-testing of items followed by applying potential items (n = 38) to a sample of of 454 students. Internal consistency (Cronbach’s alpha), Kaiser-Meyer-Olkin Criterion (KMO), Bartlett’s test, Velicer’s minimum average partial (MAP) test for number of underlying factors and Exploratory Factor Analysis using Principal Component Analysis with oblimin rotation and stepwise item reduction were assessed. Results Velicer’s MAP identified 3 factors, Cronbach’s alpha = .92, KMO = .92, Bartlett’s Test revealed significant correlation. Seven items were removed from the scale due to low factor loadings (< .3). Factor 1 (access) explains 18% of variance, factor 2 (knowledge) 12% and factor 3 (reflection) 10%, respectively. Overall, the model explains 40% of variance. Model fit represented by RMSEA index = .05. Conclusions A new general EHL scale has been developed and proven to be suitable for a students sample. This marks the first step towards a validated tool for multiple purposes. Cultural adaptation as well as multiple language validations are in process. Key messages • The process towards a general Environmental Health Scale has started. • The development of a validated EHL scale will improve understanding on whether the general population or subgroups are able to gain, understand and act upon information regarding environmental health.

Examining Retention Methods in Factor Analysis: A Comparison of Polychoric and Pearson Correlations for Categorical and Continuous Data.

In the last two decade, it has become clear that retention methods must utilize polychoric correlation instead of Pearson correlation to eliminate drawbacks such as underestimation of the magnitude of the relationship between latent variables that result in spurious findings (Bernstein & Teng, 1989). In the present study, the literature review will be examined and compared using Monte Carlo simulation to determine the most parsimonious method of retention for categorical data. With continuous variables, the majority of researchers still implement Cattell’s scree test (Henson & Roberts, 2006) and Kaiser-Guttman-1 rule (Velicer et al., 2000), because these procedures are the default in popular statistical packages, such as SPSS and SAS. The present study will examine two of the most accurate methods: MAP (Minimum Average Partial) and PA (Parallel Analysis) along with Very Simple Structure (VSS) with categorical variables.

The Rosenberg Self-esteem Scale: A Confirmatory Factor Analysis Study

<p style="text-align: justify;">The aim of the present study is to conduct a confirmatory factor analysis (CFA) of the Rosenberg Self-esteem Scale (RSES) as part of the study of affective variables using a sample of English as a foreign language (EFL) university students in Morocco. Two hundred and six (N = 206) participants of undergraduate, graduate, and post-graduate levels completed the self-esteem (SE) questionnaire. Using classical methods of factor extraction before employing more robust techniques comprising minimum average partial (MAP) and parallel analysis (PA) to perform preliminary factor analysis (FA) using principal axis factoring (PAF), results conclusively and parsimoniously yielded a one-factor solution with acceptable construct reliability (Composite Reliability). CFA results, including goodness-of-fit indexes, confirmed that the one-factor model was better fitting compared to its competing independent two-factor counterpart, but marginally less so compared to the correlated version of the latter. Two out of the three constructed models showed good fit indexes, thus demonstrating the conformity of two measurement models with their respective hypothesized structural models. Furthermore, using the heterotrait-monotrait (HTMT) ratio, both two-factor models showed acceptable discriminant validity. The obtained results further corroborate both the one-factor and two-factor solutions reported in previous works for which we present new evidence from a Moroccan EFL context.</p>

Psychometric development and validation of victim gaslighting questionnaire (VGQ): across female sample from Pakistan

PurposeGaslighting a form of abusive manipulation both emotional and psychological is a growing phenomenon in recent times. However, as of yet, there is a scarcity of a valid and reliable instrument which can measure the severity of gaslighting in victims of interpersonal relationships abuse. The purpose of this study is the development of an instrument which can effectively measure gaslighting in victims and is psychometrically reliable and valid.Design/methodology/approachSince the aim of the study was the development of a scale first a sample of eight women who were victims of domestic abuse was taken for the focus group. Afterwards using purposive sampling a sample of 20 women for the pilot study and a sample of 150 women for the main study was taken with age range 18–40 (M = 23.38, S.D = 4.03). For the development of scale theoretical basis along with a focus group was conducted to establish an item pool. Afterwards, subject matter experts helped in establishing contend validity followed by Velicer’s minimum average partial (MAP) method and maximum likelihood factor analysis (FA) was performed for the establishment of the factorial structure of the instrument.FindingsVelicer’s MAP method and Maximum Likelihood FA suggested two factor structures including peer disagreement and loss of self-trust. Instrument displayed high alpha reliability of α = 0.934, with α = 0.927 and α = 0.854, for the subscale, respectively.Research limitations/implicationsThough all necessary steps were taken to minimize the limitations of the present study, however, some limitations do exist which needs to be addressed. The foremost limitation of the present scale is that it is being developed with only a female sample, however, the inclusion of a male sample in future studies can help in identifying whether men also are victims of gaslighting from peers and other family members or not. The second limitation is of validity though necessary validities have been established future studies should study on establishing further validities to further refine the instrument. Additionally, the scale has only been validated and tested on female samples future studies should be conducted on other specific groups or samples to develop norms. Moreover, testing the scale on other cultures could also help in establishing cross cultural validation of the instrument. Finally, though the scale assumes a higher level of scores suggests a higher level of victimization, a proper cutoff score can help in further identifying proper victims from the normal level of gaslighting.Practical implicationsThe present instrument has its applicability in several domains the most important being in the criminal justice system as gaslighting comes under gaslighting and even in the UK is considered as a criminal offense. This instrument can help in determining the severity of gaslighting in victims. Likewise, it can be used in clinical settings for psychologists to identify possible cases of gaslighting victims which can enable them to provide specific help and treatment for them. Moreover, researchers can also benefit from the instrument as it can enable them to explore gaslighting with other possible variables which can help them explore the concept of gaslighting even further.Originality/valueThis paper is a novel study and has been completed with the purpose of evaluating the effects of gaslighting in victims of interpersonal relationships abuse as the earlier measures are either not psychometrically valid or cannot be generalized to a wider population. The present established scale is an effort to construct an instrument that can be used worldwide.

Cross-cultural adaptation, from Dutch to Swedish language, of the Individual Work Performance Questionnaire.

There is a need for a short, self-rated, validated and reliable instrument for individual work performance suitable for generic use in the Swedish work and organizational context. The Individual Work Performance Questionnaire (IWPQ), comprising originally 47 items, was initially developed in the Netherlands, based on a four-dimensional conceptual framework, in which individual work performance consisted of task performance, contextual performance, adaptive performance, and counterproductive work behavior. During the development process, IWPQ was shortened to 18 items with three scales formally labeled as Task performance, Contextual performance, and Counterproductive work behavior (CWB), capturing three work performance types. The current version of the IWPQ, consisting of 18 items and three scales, was then translated as well as cross-culturally adapted to American-English and Indonesian contexts. To translate and adapt the current IWPQ version, consisting of 18 items, from the Dutch to the Swedish context, to assess its content validity through cognitive interviews, to apply it to a pilot group to present descriptive statistics, to calculate the questionnaire's internal consistency, as well as to clarify whether the translated items capture three or four performance types. The Dutch version of the IWPQ, consisting of 18 items, was translated into Swedish. A six-stage translation and adaptation process was used: forward translation, synthesis, back translation, harmonization, cognitive interviews, revision, and sampling and analyses of pilot data for 206 managers (149 women) from five Swedish municipalities. IWPQ instructions, wording of a few items and one response form were slightly modified. The pilot testing showed Cronbach's alphas similar to the Dutch version of the IWPQ, ranging between 0.73 and 0.82, good mean-inter-item correlations (all above 0.36). In deciding how many factors to retain, we employed both parallel analysis (PA), and Velicer's minimum average partial (MAP) test. The number of factors to retain was, as indicated by PA, four, and by MAP, three or four. Exploratory factor analysis (principal axis factoring) revealed clearly separate factors, corresponding to four, rather than three, performance types. A new factor, roughly representing adaptive performance, comprised in the original, longer version of the IWPQ, emerged. The Swedish version of the IWPQ was successfully translated and adapted in a pilot group of managers. Before it is used, it should be validated in a larger group of managers and in more heterogeneous groups of both white- and blue-collar workers.

Academic Library Use is Positively Related to a Variety of Educational Outcomes

A Review of:
 Soria, K. M., Fransen, J., & Nackerud, S. (2017). Beyond books: The extended academic benefits of library use for first-year college students. College & Research Libraries, 78(1), 8-22. https://doi.org/10.5860/crl.78.1.8
 Abstract
 Objectives – To consider the relationship between academic library use and four specific outcomes: academic engagement, engagement in scholarly activities, academic skills development, and grade point average.
 Design – Hierarchical regression analysis.
 Setting – A large, public research university in the Midwest US.
 Subjects – 1,068 non-transfer, first-year students who voluntarily completed the Student Experience in the Research University (SERU) survey.
 Methods – The SERU survey results were analyzed alongside student data derived from institutional records and 10 library usage variables generated from library systems. Velicer’s minimum average partial (MAP) method was employed to develop a factor analysis. Hierarchical regression analyses measured the relationships between independent variables (demographic characteristics, collegiate experiences, and libraries use) and dependent variables (students’ academic engagement, academic skills, engagement in scholarship, and fall semester grade point average).
 Main Results – Students’ use of academic libraries was reported to have a positive relationship with all four dependent variables, above and beyond those explained by pre-college and collegiate experiences: academic engagement (R2∆= .130, p < 0.001), academic skills development (R2∆= .025, p < 0.001), fall semester grade point average (R2∆= .018, p < 0.001), and engagement in scholarship (R2∆= .070, p < 0.001). Use of books and web-based library resources had the most positive relationships with academic outcomes; workshop attendance and use of reference services had limited positive relationships with academic outcomes; and use of library computer workstations had no significant effects on academic outcomes.
 Conclusion – Undergraduate student use of the academic library is positively associated with diverse academic outcomes. Although the explanatory power of library use was relatively low, ranging from 1.8 to 13.0 percent of final variance in the dependent variables, library use is nonetheless reported to contribute significantly to academic outcomes.

An Asian study on clinical and psychological factors associated with personal recovery in people with psychosis

BackgroundDespite the rising recognition of personal recovery, there is a lack of research on personal recovery in individuals with psychosis in Singapore. This study aims to evaluate the psychometric properties of the QPR-15 using the CHIME personal recovery framework and to examine its associations with clinical recovery factors.MethodsSixty-six stable outpatients were recruited and assessed at two time points approximately 2 weeks apart. Convergent validity was examined through Spearman correlations with scores on CHIME-related psychological factors: connectedness (Ryff subscale- positive relations with others), hope (Herth Hope Index- abbreviated), identity (Ryff subscale- self-acceptance, Internalized Stigma of Mental Illness- Brief), meaning (World Health Organization Quality of Life Assessment-Brief Form), empowerment (Empowerment Scale). Pearson’s correlation was used to examine the test-retest reliability, while Cronbach’s alpha was used to examine internal consistency. The initial factor structure was evaluated via principal component analysis, Velicer’s minimum average partial (MAP) criteria, parallel analysis, and a scree plot. Spearman correlations and hierarchical multiple linear regression (controlling for age and gender) were employed to examine the association of clinical (symptoms and functioning) and psychological factors with the QPR-15.ResultsThe QPR-15 demonstrated convergent validity with all CHIME-related psychological factors (rs ranged from 0.472 to 0.687). Internal consistency was excellent (Cronbach’s alpha = 0.934), and test-retest reliability was adequate (r = 0.708). Initial factor structure evaluations revealed a one-factor model. Correlations of clinical factors with the QPR-15 were mostly low (rs ranged from − 0.105 to − 0.544) but significant, except for depressive symptoms (CDSS: rs = − 0.529 to − 0.544), while correlations were moderate for psychological factors. Clinical factors significantly explained 28.3–31.8% of the variance of the QPR-15. Adding psychological factors significantly increased the model variance at baseline (∆ adjusted R2 = 0.369, F change < 0.001) and at time point 2 (∆ adjusted R2 = 0.208, F change < 0.001).ConclusionOur results provide preliminary evidence that the QPR-15 has adequate psychometric properties in Singapore and encompasses the CHIME personal recovery framework. In addition, our results suggest that clinical recovery and personal recovery are not substitutes for each other but rather are complementary, thereby promoting a more holistic evaluation of recovery in people with psychosis. Implications are discussed.

Factor structure and test-retest reliability of the Polish version of the Clarity of Auditory Imagery Scale

Vividness of imagery usually refers to the degree of similarity between mental images and corresponding percepts of real objects. One of the recently developed questionnaires, proposed to measure the vividness of auditory imagery, is the Clarity of Auditory Imagery Scale (CAIS). The main goal of the present study was to assess the factor structure, internal consistency, and test–retest reliability of the Polish version of the CAIS. The study was conducted on musicians (N = 39) and non-musicians (N = 40) to establish differences between the two groups in the vividness (or more specifically, clarity) of their auditory images. A combination of the minimum average partial (MAP) test and parallel analysis (PA) was used as a method of establishing the number of factors and provided evidence that the CAIS is one factor questionnaire. Test–retest reliability was measured by the intraclass correlation coefficient (ICC) between the mean scores obtained in two measurements made over a one-week interval. The test–retest (ICC) obtained between two measurements equaled .85. The ICC value showed satisfactory stability of the measurement of the vividness of auditory images, at least for short time intervals. The internal consistency of the scale was also satisfactory (Cronbach’s α = .87). Summarizing, the psychometric properties of the Polish version of the CAIS indicate that the scale is a reliable measure of the vividness of auditory imagery. Vividness of auditory imagery measured by the CAIS was not influenced by sex or musical expertise factors.

The structure and function of pathogen disgust.

Researchers have long noted that many of the multiple elicitors of disgust have some relation to infectious disease. There is an emerging consensus that disgust evolved in Animalia to direct the behaviours that reduce risk of infection, so-called 'parasite avoidance theory'. If this is correct, then the disgust motive should be structured in a manner that reflects the ways in which infectious disease can be avoided. In this study, we generated a set of items based on the epidemiology of disease transmission. These were then rated for their capacity to elicit disgust by a large, predominantly North American/UK sample and subjected to factor analysis to identify latent variables. While a number of plausible factor solutions emerged, Velicer's MAP (minimum average partial) test suggested six domains: atypical appearance, lesions, sex, hygiene, food and animals. This structure did not exactly mirror the transmission routes of infections, as we initially predicted, but it may rather reflect distinct kinds of behavioural tasks involved in avoiding disease. This finding makes sense from the perspective of a cognitive system that evolved under selection for a behavioural response to threats from the social and biological environment. We suggest that regularly occurring types of infectious disease problems have produced regularities in the domain structure of pathogen disgust and discuss the implications of these results for understanding the structure, function and measurement of motives such as disgust in humans and other animals.This article is part of the Theo Murphy meeting issue 'Evolution of pathogen and parasite avoidance behaviours'.

Validation of the revised piper fatigue scale in Koreans with chronic hepatitis B.

PurposeThe purpose of the study was to evaluate the construct validity and reliability of the Korean version of the revised Piper Fatigue Scale (PFS) in Koreans with chronic hepatitis B.MethodsA total of 146 chronic hepatitis B patients completed the Korean version of the revised PFS. A descriptive analysis was performed to determine the subjects’ demographic characteristics; the construct validity was examined using exploratory factor analysis; and internal consistency reliability of the scale was estimated for the meaningful total scale and factors.ResultsThe factor analysis supported the original four-factor structure based on Kaiser Criterion and Minimum Average Partial (MAP): Behavioral/Severity, Affective meaning, Sensory, & Cognitive/Mood. In the 22 items in the original instrument, patient/impatient, relaxed/tense, and exhilarated/depressed were re-identified from the cognitive/mood subscale and sensory subscale. The Cronbach’s alpha of the 22-item Korean version of the revised PFS was 0.96 for the total scale, and the range of Cronbach’s alpha for subscales was 0.90 to 0.93.ConclusionsThe results of the study revealed that the 22-item Korean version of the revised PFS is valid and reliable in Koreans with chronic hepatitis B. Further studies ascertaining the psychometric properties of the revised PFS need to be performed in Korean patients.

Validation and application of the Chinese version of the 10-item Connor-Davidson Resilience Scale (CD-RISC-10) among parents of children with cancer diagnosis.

Parents of children diagnosed with cancer often experience considerable emotional distress for their children with negative emotions, such as disbelief, depression, anxiety, hope and shock. Resilience is defined as the psychological characteristics that promote positive adaptation in the face of stress and adversity, which has been demonstrated to relate to positive coping and less psychological distress. Thus, a quick screening tool to evaluate the levels of resilience of parents with cancer-diagnosed children is urgently required. The Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) were used to evaluate the CD-RISC-10 using a sample size of 500 parents. Velicer's Minimum Average Partial (MAP) Test and a parallel analysis were also supplemented to confirm the EFA-derived structure of the scale. The participants were given the 10-item Kessler Psychological Distress Scale (K-10), Perceived Social Support Scale (PSSS) and Medical Coping Modes Questionnaire (MCMQ) to test the associates with CD-RISC-10 and obtain the cut-off of the scale. The Chinese version of CD-RISC-10 has good psychometric properties and retains its single dimension in the original English version, which can explain 49.602% of the total variance. The CFA demonstrates the fit indices of a one-order model: Chi-Square=39.987, CMIN/DF=1.333, P<0.001, TLI=0.914, CFI=0.981, GFI=0.962, NFI=0.926, IFI=0.979, RFI=0.889, RMR=0.042, and RMSEA=0.041. The CD-RISC presents statistical associations with other scales, and the cut-off is 25.5. The Chinese version of the CD-RISC-10, which is reliable, valid and easy to use, is suitable for clinical settings. The CD-RISC-10 enables a quick understanding of the level of resilience of the parents when their children undergo treatment, which can be the most important indicator to their psychological health.

Dopamine Receptors and the Pharmacogenetics of Side-Effects of Stimulant Treatment for Attention-Deficit/Hyperactivity Disorder

There is considerable interest in the pharmacogenetics of responses to stimulant treatment for attention-deficit/ hyperactivity disorder (ADHD) (Froehlich et al. 2010). By contrast, little attention has been paid to individual differences in side effects. Dopamine function in the prefrontal cortex (PFC) is a strong candidate system for understanding these (Arnsten, 2009). D1 receptors are the most abundant dopamine (DA) receptor subtype, and mediate most of the cellular effects of DA in this region (Lewis and Gonzalez-Burgos 2006). They are thought to be involved in the regulation of sustained firing of dorsolateral PFC neurons during the delay phase of delayed-response tasks that require working memory (Sawaguchi and Goldman-Rakic 1994; Seamans and Yang 2004; Lewis and Gonzalez-Burgos 2006). Sawaguchi and Goldman-Rakic (1994) showed that D1 dopamine receptors in the dorsolateral prefrontal cortex of monkeys participated in the maintenance of internalized visuospatial representations and/or in the control of eye movements governed by internal cues. They suggested that the D1 family of dopamine receptors may have a critical role in PFC-mediated working memory and cognitive deficits. Variations in D1 polymorphism rs4532 are associated with ADHD (Bobb et al. 2005) and predict side effects to medication for schizophrenia (Potkin et al. 2003; Lai et al. 2011), although in the latter studies the effects may relate to excess dopamine at the D1 receptor. In both cases, there are good reasons to suggest that D1 receptor genetics may predict side-effects from stimulant medication for ADHD. We tested the relationship of stimulant side effects to common variations in DA receptor genes using 90 Caucasian children (79.2% males) who met formal criteria for American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV) diagnosis of ADHD (American Psychiatric Association 1994); were 4–14 years of age (mean = 8.85 years; SD= 2.32); had no major neurological/physical illness; had an intelligence quotient (IQ) >70; were using stimulant medication; and were not using another psychoactive medication (full demographic and diagnostic data and medication use available from Dr. Levy). Ethics approval and fully informed consent was from the University of New South Wales. Participants gave blood (52.5%) or saliva (or both for a smaller reliability check sample) via Oragene saliva collection kits (http://www.dnagenotek.com/). DNA extraction rates were >95% for both methods. Samples were genotyped for four dopamine receptor single nucleotide polymorphisms (SNPs): DRD1 rs4532, DRD2 rs6277, DRD3 rs6280; DRD4 rs7932167, and COMT rs4680. All were in Hardy–Weinberg equilibrium. Genotypes were determined using iPLEX Gold primer extension followed by mass spectrometry analysis on the Sequenom MassARRAY system (Sequenom, San Diego, CA) by the Australian Genome Research Facility (AGRF: http://www.agrf.org .au/). Age, weight sex, and IQ were noted, and the side effects rated as absent, mild, moderate, and severe, using a modified Barkley side-effect scale (Barkley et al. 1990). Medication usage was as follows: Primary medication Ritalin (immediate release [IR]) n = 43, dosage mean = 22.03 SD = 11.7; Ritalin (modified release) n = 14, dosage mean = 26.43 SD = 6.33; Concerta n = 12, dosage mean = 29.25 SD = 7.79; dexamphetamine n = 19, dosage mean = 8.37 SD = 4.47; dexamphetamine (long acting [Dex LA]) n = 2, dosage mean = 10 SD = 0; secondary medication Ritalin (IR) n = 5, dosage mean = 14.0 SD = 10.25; Catapres n = 9, dosage mean = 25.05 SD = 10.83. Secondary medications were distributed as follows: 4 of the 43 children receiving primary Ritalin IR were also receiving Catapres at night; 6 of the 14 children receiving Ritalin (modified release) were receiving a secondary medication, that is, 3 were also receiving Catapres at night, and 3 were receiving an additional Ritalin IR during the day. Two of the 19 children receiving dexamphetamine were also receiving Ritalin IR during the day; none of those on Dex LA were receiving an secondary medication; 2 of the 12 on Concerta were receiving Catapres at night. Medication doses were standardized to a standard Ritalin (IR) dose of 10 mg by the following formula: Dexamphetamine 5mg =Ritalin IR 10mg; Ritalin (modified release) 10mg =Ritalin IR 5mg b.i.d.; Concerta 18mg =Ritalin IR 5mg t.i.d. We reduced the 16 side effects items to factors using principal components analysis. Both parallel analysis and Velicer’s minimum average partial (MAP) test indicated that there were three nonrandom factors; therefore, we generated a three factor solution using Varimax rotation with Kaiser normalization. Factor loadings are available from Dr. Levy. The three factors, nausea, withdrawal

Performance of Velicer’s Minimum Average Partial Factor Retention Method With Categorical Variables

Despite strong evidence supporting the use of Velicer’s minimum average partial (MAP) method to establish the dimensionality of continuous variables, little is known about its performance with categorical data. Seeking to fill this void, the current study takes an in-depth look at the performance of the MAP procedure in the presence of ordinal-level measurement. Using Monte Carlo methods, seven factors related to the data (sample size, factor loading, number of variables per factor, number of factors, factor correlation, number of response categories, and skewness) as well as two factors related to the MAP method (type of correlation matrix and power) were systematically manipulated. The results indicate that using polychoric correlations and the squared partial correlations leads to considerably more accurate estimations than using Pearson correlations and/or raising the partial correlations to the fourth power. Additionally, the MAP method is shown to be a biased estimator of dimensionality in two conditions: (a) for low factor loadings (.40) and (b) for medium factor loadings (.55) and a small number of variables per factor (≤ 6). The applicability of this method with categorical variables is discussed in the context of these findings.

Factors perceived to influence the selection of information technology jobs

There are huge tangible as well as intangible costs involved in the recruitment process, especially for information technology (IT) companies. Reduced recruitment costs would indirectly boost the business performance. The present work is a study undertaken to assess the factors perceived to influence the selection of the IT jobs using factor analysis for three IT companies operational in India differing in certain parameters. This study uses two highly promising but underutilised factor retention criteria namely, parallel analysis (PA) and minimum average partial (MAP) technique. We found that these two retention criteria provide consistent results in determining the number of factors and the use of PA and MAP in factor analysis can significantly reduce the subjective interpretation of supposedly objective methods. The study would be useful for the human resource department of the IT firms to understand the psyche of the candidates before making a recruitment pitch in front of them.

Factor Analysis Revisited – How Many Factors are There?

We suggest a procedure to improve the overall performances of several existing methods for determining the number of factors in factor analysis by using alternative measures of correlation: Pearson's, Spearman's, Gini's, and a robust estimator of the covariance matrix (MCD). We examine the effect of the choice of the covariance used on the number of factors chosen by the KG rule of one, the 80% rule, the Minimum average partial (MAP), and the Parallel Analysis Methodology (PAM). Extensive simulations show that when the entire (or part) of the data come from heavy-tail (lognormal) distributions, ranking the variables which come from non symmetric distributions improves the performances of the methods. In this case, Gini is slightly better than Spearman. The PAM and MAP procedures are qualitatively superior to the KG and the 80% rules in determining the true number of factors. A real example involving data on document authorship is analyzed.

Using exploratory factor analysis in personality research: Best-practice recommendations

Orientation: Exploratory factor analysis is the method of choice with objective personality instruments, particularly to establish the construct validity and construct equivalence of trait-based instruments.Research purpose: This article presents more objective methods to determine the number of factors, most notably parallel analysis and Velicer’s minimum average partial (MAP). The benefits of rotation are also discussed. The article argues for more consistent use of Procrustes rotation and congruence coefficients in factor analytic studies.Motivation for the study: Exploratory factor analysis is often criticised for not being rigorous and objective enough in terms of the methods used to determine the number of factors, the rotations to be used and ultimately the validity of the factor structure.Research design, approach and method: The article adopts a theoretical stance to discuss the best-practice recommendations for factor analytic research in the field of psychology. Following this, an example located within personality assessment and using the NEO-PI-R specifically is presented. A total of 425 students at the University of the Witwatersrand completed the NEO-PI-R. These responses were subjected to a principal components analysis using varimax rotation. The rotated solution was subjected to a Procrustes rotation with Costa and McCrae’s (1992) matrix as the target matrix. Congruence coefficients were also computed.Main findings: The example indicates the use of the methods recommended in the article and demonstrates an objective way of determining the number of factors. It also provides an example of Procrustes rotation with coefficients of agreement as an indication of how factor analytic results may be presented more rigorously in local research.Practical/managerial implications: It is hoped that the recommendations in this article will have best-practice implications for both researchers and practitioners in the field who employ factor analysis regularly.Contribution/value-add: This article will prove useful to all researchers employing factor analysis and has the potential to set the trend for better use of factor analysis in the South African context.

Higher-order exploratory factor analysis of the Reynolds Intellectual Assessment Scales with a referred sample

The factor structure of the Reynolds Intellectual Assessment Scales (RIAS; [Reynolds, C.R., & Kamphaus, R.W. (2003). Reynolds Intellectual Assessment Scales. Lutz, FL: Psychological Assessment Resources, Inc.]) was investigated with a large ( N = 1163) independent sample of referred students (ages 6–18). More rigorous factor extraction criteria (viz., Horn's parallel analysis (HPA); [Horn, J.L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179–185.], and Minimum Average Partial (MAP) analysis; [Velicer, W.F. (1976). Determining the number of components from the matrix of partial correlations. Psychometrika, 41, 321–327.]), in addition to those used in RIAS development, were investigated. Exploratory factor analyses using both orthogonal and oblique rotations and higher-order exploratory factor analyses using the Schmid and Leiman [Schmid, J., and Leiman, J.M. (1957). The development of hierarchical factor solutions. Psychometrika, 22, 53–61.] procedure were conducted. All factor extraction criteria indicated extraction of only one factor. Oblique rotations resulted in different results than orthogonal rotations, and higher-order factor analysis indicated the largest amount of variance was accounted for by the general intelligence factor. The proposed three-factor solution was not supported. Implications for the use of the RIAS with similarly referred students are discussed.

Historical increase in the number of factors measured by commercial tests of cognitive ability: Are we overfactoring?

A historical increase in the number of factors purportedly measured by commercial tests of cognitive ability may result from four distinct pressures including: increasingly complex models of intelligence, test publishers' desires to provide clinically useful assessment instruments with greater interpretive value, test publishers' desires to include minor factors that may be of interest to researchers (but are not clinically useful), and liberal statistical criteria for determining the factor structure of tests. The present study examined the number of factors measured by several historically relevant and currently employed commercial tests of cognitive abilities using statistical criteria derived from principal components analyses, and exploratory and confirmatory factor analyses. Two infrequently used statistical criteria, that have been shown to accurately recover the number of factors in a data set, Horn's parallel analysis (HPA) and Minimum Average Partial (MAP) analysis, served as gold-standard criteria. As expected, there were significant increases over time in the number of factors purportedly measured by cognitive ability tests ( r = .56, p = .030). Results also indicated significant recent increases in the overfactoring of cognitive ability tests. Developers of future cognitive assessment batteries may wish to increase the lengths of the batteries in order to more adequately measure additional factors. Alternatively, clinicians interested in briefer assessment strategies may benefit from short batteries that reliably assess general intellectual ability.

SPSS and SAS programs for determining the number of components using parallel analysis and velicer's MAP test.

Popular statistical software packages do not have the proper procedures for determining the number of components in factor and principal components analyses. Parallel analysis and Velicer's minimum average partial (MAP) test are validated procedures, recommended widely by statisticians. However, many researchers continue to use alternative, simpler, but flawed procedures, such as the eigenvalues-greater-than-one rule. Use of the proper procedures might be increased if these procedures could be conducted within familiar software environments. This paper describes brief and efficient programs for using SPSS and SAS to conduct parallel analyses and the MAP test.

Robustness is Not Dimensionality: On the Sensitivity of Component Comparability Coefficients to Sample Size.

Everett (1983) has proposed that, under certain conditions, replicability provides an answer to the question of the number of dimensions to retain in component analysis. But replicability must logically be a function of sample size as well as dimensionality. In the present study, the effects of sample size and sample composition are systematically examined on the replicability of principal components. Using observer ratings of personality from the California Adult Q-Set, comparability coefficients are examined in 192 series of principal components analyses. Results indicate that (a) once one has 20 or more subjects per item, the most comparable solution typically has as many components as items; (b) if these full solutions are ignored, there is still a substantial relationship between prescribed number of components and sample size when one uses either the .90 or .85 threshold decision rules and (c) other criteria for determining the number of components to retain, such as the Minimum Average Partial (MAP) rule, do not show the same relationship with sample size. These results indicate that dimensionality cannot be inferred from component robustness, as these are empirically as well as logically separate matters.