No abstract

<span style="font-family: 'Garamond',serif; font-size: 8pt; mso-bidi-font-size: 10.0pt; mso-fareast-font-family: 'Times New Roman'; mso-bidi-font-family: 'Times New Roman'; mso-ansi-language: EN-US; mso-fareast-language: AR-SA; mso-bidi-language: AR-SA;" lang="EN-US">The aim of the present study is to illustrate how the appropriate or inappropriate application of exploratory factor analysis (EFA) can lead to quite different conclusions. To reach this goal, we evaluated the degree to which four different programs used to perform an EFA, specifically SPSS, FACTOR, PRELIS and MPlus, allow or limit the application of the currently recommended standards. In addition, we analyze and compare the results offered by the four programs when factor analyzing empirical data from scales that fit the assumptions of the classic linear EFA modeling adequately, ambiguously, or optimally, depending on the case, through the possibilities the different programs offer. The results of the comparison show the consequences of choosing one program or another; and the consequences of selecting some options or others within the same program, depending on the nature of the data. Finally, the study offers practical recommendations for applied researchers with a methodological orientation.</span>

The objective of this study is to test the validity of the collective climate concept. It was expected that membership in collective climates was related to membership in the collectivities de ned by departmental membership, hierarchical level, shift, job location and organizational tenure. The study sample was composed of 195 employees from a central administration agency. Using a combination of hierarchical and non-hierarchical clustering methods, three diVerent collective climates were obtained. The results showed that only hierarchical level was related to collective climate membership. Based on all the results obtained, the debate on the validity of collective climates is reconsidered, and research on climate formation and relevant factors in that process is emphasiz ed. V icente González-Romá et al.

La validit6 de construction des Cchelles de Rizzo, House, et Lirtzman (1970) a CtC fortement critiquCe ces dernikres annCes. Les critiques portent sur la formulation des items qui constituent ces Cchelles, ce qui fait que l'interprttation des Cchelles dCpend totalement de la formulation des items. Ces critiques ont favoris6 1'Cmergence d'autres modkles factoriels. La duplication tlargie de Kelloway et Barling (1990), l'analyse factorielle destinte a confirmer la validit6 des Cchelles de Rizzo et al. et une Cchelle de surcharge des r8les avec des items formulCs a la fois affirmativement et nigativement furent utiliskes pour Cprouver la validitC de construction des tchelles d'ambiguitb et de conflit de r8les. Cinq modkles factoriels possibles furent comparCs sur deux Cchantillons (N, = 173 et N, = 273). Un modkle qui propose trois facteurs de premier ordre compatibles avec des interprktations indkpendantes des conflits de r61e et de la surcharge des rbles, et un facteur de second ordre concernant le stress des r6les en gCnCral correspondait mieux aux donnCes que les quatre autres modkles dans les deux Cchantillons, confirmant ainsi la validitt de construction des tchelles.The construct validity of Rizzo, House, and Lirtzman's (1970) scales has been seriously criticised during recent years. The criticisms are based on the wording of the items comprising these scales, which causes the substantive interpretation of the scales to be perfectly confounded with the direction of item wording. The criticisms of the scales have stimulated alternative factor models. Replicating and extending Kelloway and Barling's (1990) study, confirmatory factor analysis of Rizzo et al.'s scales and a role overload scale consisting of positively and negatively worded items was used to test the construct validity of role conflict and ambiguity scales. Five alternative factor models were compared in two distinct samples (N, = 173 and N, = 273). A model that posits three first-order factors consistent with substantive interpretations of role conflict, role ambiguity, and role overload, and a Requests for reprints should be sent to Dr. The authors wish to thank three anonymous reviewers and the editor for their helpful comments on an earlier draft. 0 1998 International Association of Applied Psychology 536 GONZALEZ-ROMA AND LLORETsecond-order general role stress factor, fitted the data better than the other four models in both samples, thus supporting the scales' construct vaiidity.

No abstract