Letter to Editor: Parallel analysis and MBI-HSS: How many factors

 Products & Services

of 3
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Letter to Editor: Parallel analysis and MBI-HSS: How many factors
  247  Merino C   /Colombia Médica - Vol. 44 Nº 4 2013 (Sep-Dec) Colombia Médica colombiamedica.univalle.edu.co Colombia Médica Journalhomepage: http://colombiamedica.univalle.edu.co Facultad de SaludUniversidad del Valle Letter to Editor Parallel analysis and MBI-HSS: How many factors? Mr. Editor:  It has been only recently possible to validate the  Maslach Burnout Inventory-Human Services Survey   (MBI-HSS) 1 among health proessionals o Cali 2 , an important step or using this instrument with local empirical support in regard to its re-liability o scoring and internal structure. However, two aspects o this analysis can be considered as methodological weaknesses. First, the Cronbach alpha coefficient was calculated or the total group o items, and this is absolutely inappropriate because: a)the authors did not demonstrate empirical support or accomplishing this (e.g., a hierarchical actor analysis), b) the literature indicates that actors in the MBI-HSS are generally independent, a char-acteristic also reported by Córdoba et al  .  2 ,   and the same authors o the MBI-HSS 1 ) the authors did not report the inter-actor cor-relations with which an appreciation could have been obtained, at least an heuristic one o the common degree o variance among the actors.Secondly, the authors obtained seven actors in their exploratory actor analysis; this large number off actors seems to be a product o applying a actor retention method that is now consensually seen as inaccurate and little recommended 3,4. Spe-cifically, it is known as Kaiser’s rule , Guttman´s rule  or simply  K1 4 . Te problem identified with this method is its over-estimation o the number o actors to be retained 3,4 , a situation that clearly oc-curs in the results reported by Cordoba et al  . 2 , as reported in their able 2.A more accurate method which has gained a scientific consensus or good practices or retaining the number o actors is called  parallel analysis 3,4 . Tis procedure is based on the work o Horn 4 , which consists o randomly creating the same number o vari-ables as the number o items analyzed (in the case o MBI-HSS, 22 items),correlating them and extracting eigenvalues against which the eigenvalues derived rom the empirical data under analysis are compared. Tis procedure was applied to the eigenvalues reported by Cordoba et al. 2 , in able 2 by means o the  Monte Carlo PCA 5 program (100 replications). Our results are shown in able 1.Te appropriate number o actors to retain is achieved by making a one to one comparisons o the eigenvalues, keeping the empiri-cal eigenvalue that is less than the random eigenvalue. Conceptu-ally, this indicates that the significant eigenvalues must be greater than those generated randomly. In able 1, the number o eigen- values to retain is 3, which is exactly the same number o actors underlying the MBI-HSS. Validating the result with a conceptual analysis o these three actors, would leave one to conclude that the actorial structure o the MBI-HSS is replicable in the sample studied. Compared with the first result o the authors (7 actors), the methodological difference is clear.Finally, we note that the use o parallel analysis should be recommended or making more ac-curate decisions about the number o actors to retain. References 1. Maslach C, Jackson, S. Maslach Burnout Inventory-Human Ser- vices Survey (MBI-HSS). In:   Maslach C, Jackson, S. Leiter M. editors. Maslach Burnout inventory manual. Mountain View: Consulting Ps-chologists Press; 1996.2. Córdoba L, amayo J, González M, Martinez M, Rosales A, Bar-bato S. Adaptation and validation o the Maslach Burnout inventory-human services survey in Cali, Colombia. Colomb Med. 2011; 42(3): 286-93.3. Hayton JC, Allen DG, Scarpello V. Factor retention decisions in exploratory actor analysis: A tutorial on parallel analysis. Organ Res Meth. 2004; 7(2): 191-205.4. Dinno A. Exploring the sensitivity o Horn’s parallel analysis to the distributional orm o random data. Multivariate Behav Res. 2009; 44(3): 362–88. 5.Watkins MW. Monte Carlo PCA or parallel analysis [computer sofware]. State College, PA: Ed & Psych Associates. 2000. Cesar Merino Soto,    Marisol Angulo Ramos Table 1. Te eigenvalues o Cordoba et al and those generated randomly (100 replications) No. of eigen values Córdoba et al  . Random1 4.17 1.512 2.06 1.423 1.45 1.354 1.23 1.295 1.15 1.256 1.07 1.27 1.05 1.168 - 1.119 - 1.0710 - 1.0311 - 112 - 0.9613 - 0.9314 - 0.8915 - 0.8516 - 0.8217 - 0.7918 - 0.7519 - 0.7220 - 0.6821 - 0.6422 - 0.59  248 Responding to Regarding the article, entitled “Adaptation and validation o the Maslach Burnout Inventory-Human Services Survey in Cali, Colombia” as published in Colombia Médica, please note the o-llowing clarifications:Cronbach´s Alpha Coefficient or the total MBI scale can be con-sidered inappropriate in the usage and evaluation o the MBI gi- ven the independence o the actors or the MBI-HSS; however, we, the authors do not share this statement as this result does not affect the conclusions o the study. .An exploratory actor analysis was used mainly as a method to cross-validate the item analysis previously conducted 1  and, se-condarily, to examine the structure o relationships between va-riables, to detect possible multi-dimensionality o the construct assessed, and to explore the validity o the construct or the MBI-HSS so that the underlying dimensions o the items in the context could be identified 2 . From this analysis, the researchers wanted to explore the internal structure and dimensionality proposed in the theoretical model or the MBI-HSS rom the data collected without making assumptions about the same model with three dimensions evaluated in the context; as the title suggests, its pur-pose was exploratory in nature 1, 2, 3 . However, later the confirma-tory actor analysis was used in order to statistically contrast the hypothesis based on the grouping o the items proposed by the theory or model suggested by the MBI-HSS authors. Tis analysis allowed the researchers to test the hypothesis by inerential tech-niques and provide inormative analytical options 4 . Currently, it is recommended that beore proceeding with the application o a confirmatory actor analysis that exploratory actor analysis pro-cedures are used 1 , as was previously mentioned.For the exploratory actor analysis, Kaiser´s criterion was used as a actor retention method, although today it is not the method most recommended 6,7,8 ; in numerous articles it has been used as a tool to obtain a first approximation o the actorial structure o the MBI-HSS 9,10,11 . As the more variables in the analysis the less the  variance needed to explain a actor, so the Kaiser criterion tends to suggest too many actors 7 . Tereore, some suggest that it be used with other indicators 12  or that a confirmatory actor analysis is conducted to validate the number o actors 5 . Currently, parallel analysis is widely accepted to determine the number o actors to be retained 5.7 .Similarly, although the exploratory actor analysis has provided a seven-actor structure, studies that have evaluated the psychome-tric properties o the MBI with other populations show a grea-ter number o actors than the srcinal version 13, 14, 15  by this same method.Reerencing the application o parallel analysis in the retention o the number o actors in the exploratory actor analysis (see Fi-gure 1), it clearly shows that the results are consistent with the confirmatory actor analysis and show strong evidence in support o the three-actor structure o the srcinal model (values grea-ter than simulated Eigen values and those rom re-sampling). It must be noted that the confirmatory model allows evaluation o the statistical fit between the srcinal MBI model and our data, and the exploratory actor analysis can in no way be used as this model incorporates ew substantive assumptions and allows or each item to depend on all common actors so that the interpreta-tion is heuristic and difficult 16 . Tereore, the confirmatory actor analysis model corrects the inherent deficiencies in the explora-tory perspective and leads to greater support or the hypothesis o the srcinal structural model o the MBI-HSS.We emphasize that through the proposed analysis one does not reject the hypothesis o a three-actor structure or evaluating Bur-nout syndrome in the population studied. Finally, it is appropriate to note that the researchers did not make any decision on the di-mensionality o MBI-HSS through the exploratory actor analysis, but rather used the confirmatory actor analysis to decide on the actorial structure o the MBI. Tis analysis was done or a acto-rial structure o seven, six, five, our and two; however, these data were not published because a good fit o the structural equation model to a actor structure as previously ound. References 1 Floyd, F.J. y Widaman, K.F. (1995). Factor analysis in the deve-lopment and refinement o clinical assessment instruments. Psy-chological Assessment, 7, 286-299.2 Hwan, S., Lee, M. (2009). Examining the psychometric pro-perties o the Maslach Burnout Inventorywith a sample o child protective service workers in Korea. Children and Youth Services Review 31, 206–210.3 Henson RK, Roberts JK. (2006). Use o Exploratory Factor Analysis in Published Research:Common Errors and Some Com-ment on Improved Practice. Educational and Psychological Mea-surement;66, 3. Figure 1. Eigen values and parallel analysis over 1000 replica-tions using the R ( 17 ) statistical sofware  Merino C   /Colombia Médica - Vol. 44 Nº 4 2013 (Sep-Dec)  2494 Tompson B. (2004). Exploratory and confirmatory actor analysis: understanding conceptsand applications. Washington, DC: American Psychological Association.5 Costello, Anna B. & Jason Osborne (2005). Best practices in ex-ploratory actor analysis: ourrecommendations or getting the most rom your analysis. Practical Assessment Research & Eva-luation, 10(7).6 Patil, V., Singh, S., Mishra, S. and Donavan, . (2008). Efficient theory development and actor retention criteria: Abandon the ‘ei-genvalue greater than one’ criterion. Journal o Business Research 61, 162–170.7 Hayton, J., Allen, D. and Scarpello, V. (2004). Factor Retention Decisions inExploratory Factor Analysis:A utorial on Parallel Analysis. Organizational Research Methods, Vol. 7 No. 2, 191-205.8 Williams, B., Brown, ., Onsman, A. (2010). Exploratory actor analysis: A five-step guide ornovices Journal o Emergency Pri-mary Health Care (JEPHC), Vol. 8, Issue 3.9 Galanakis, M., Moraitou, M., Garivaldis, F., Stalikas, A. (2009). Factorial Structure and Psychometric Properties o the Maslach Burnout Inventory (MBI) in Greek Midwives. Europe’s Journal o Psychology, pp. 52-70.10 Mojsa, J., Dylag, A., Palczynska, E. (2006). Psychometric pro-perties o a Polish version o the Maslach Burnout Inventory Ge-neral Survey (MBI-GS) in a group o Inormation and Communi-cation echnology(IC) specialists. Ergonomia IJE&HF, Vol. 28, No. 4, 351–361.11 Millán, A., DAuberterre, M. (2012). Propiedades psicométricas del Maslach Burnout Inventory-GS en una muestra multiocupa-cional venezolana. Revista de Psicología, vol.30, n.112 Nunnally, J. & Bernstein, I. (1995). eoría psicométrica. Méxi-co: McGraw-Hill.13 Olivares, Victor (2009). Analysis o Psychometric Properties o the Maslach Burnout Inventory Human Services (MBI-HSS) in Chilean Proessionals. Cienc rab. Oct-Dic; 11 (34): 217-221).14 Chao, S. F., McCallion, P. P., & Nickle, . . (2011). Factorial  validity and consistency o the Maslach Burnout Inventory among staff working with persons with intellectual disability and demen-tia. Journal O Intellectual Disability Research, 55(5), 529-536.15 Gil-Monte, P. (2002). Validez actorial de la adaptación al espa-ñol del Maslach Burnout inventory-general survey. Salud Publica Mex 2002;44:33-4016 Batista, J., Coenders, G., y Alonso, J. (2004). Análisis actorial confirmatorio. Su utilidad en la validación de cuestionarios rela-cionados con la salud. Med Clin (Barc);122(Supl 1):21-7.17 R Core eam (2013). R: A language and environment or statis-tical computing.R Foundation or Statistical Computing, Vienna, Austria. URLhttp://www.R-project.org/
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks