Development and Validation of an International English Big-Five Mini-Markers

Abstract

Examination of Saucier (1994) big-five Mini-Markers using a multinational sample (N = 491) found its psychometric properties suboptimal. Using further multinational samples, through a qualitative study (N = 23) appraising items and then a series of quantitative development and validation studies (total N = 3,068), a revised marker set was derived. This new International English Mini-Markers (a) produced better factor structures, higher scale internal consistency reliabilities, and greater orthogonality than the original set of items, (b) prove to have temporal stability, and (c) acceptable convergent validity.

Introduction

Personality dimensions have long attracted research attention (Eysenck, 1991). The five-factor model of personality (Goldberg, 1990, McCrae and Costa, 1987) has met with particularly wide application in the personality field, plus across disciplines as diverse as aviation (Grant et al., 2007), politics (Schoen & Schumann, 2007), and entrepreneurship (Zhao & Seibert 2006). The model’s wide disciplinary application has prompted the development of several brief big-five measures comprising fewer than 50 single-item adjectives or only a very few statement-based items specifically for use in applied research settings where respondent time or instrument space are constrained (e.g. Gosling et al., 2003, Langford, 2003, Saucier, 1994, Woods and Hampson, 2005).

While, as Church (2001) notes, some debate has existed about the uniform replicability of the five-factor model in emically developed measures across some cultures such as China (Cheung et al., 2001), the model’s general cross-cultural applicability has, nevertheless, stimulated translations of relatively long big-five measures into Chinese (Trull & Geary, 1997), Croatian (Mlacic & Goldberg, 2007), Italian (Terracciano, 2003), Spanish (Garcia, Aluja, & Garcia, 2004), and other languages. Such translations tend to support the broad cross-cultural applicability of the five-factor model. For example, Hendriks, Hofstee, and De Raad (1999) developed interactively in Dutch, English and German a 100-statement big-five measure that confirmed the five-factor structure in these languages and 11 others into which it was subsequently translated, including Chinese and Japanese. More recently, Schmitt, Allik, McCrae, and Benet-Martínez (2007) found robust support for the five-factor model across world regions in a study covering 56 nations using a 44-statement (173-word) big-five measure. Some brief big-five measures have also been translated, although as yet only into German (Muck et al., 2007, Rammstedt and John, 2007) and Swedish (Hochwalder, 2006).

The relatively low number of languages into which any single brief big-five measure has so far been translated has meant that some forms of applied cross-cultural research demanding brief measures have been constrained. Most particularly, applied research in cross-cultural settings where multiple national backgrounds can (a) be anticipated, are (b) of specific interest, but (c) cannot necessarily be known precisely in advance is currently impossible, except, as Thompson (2007) suggests, using English measures. Such research settings include international government bodies and firms, plus, of course, many educational institutions, where populations increasingly comprise individuals from numerous countries who, despite not always being native English-speakers, are obliged to operate organizationally in English.

Because English big-five measures have been emically developed, predominantly among native English-speakers in North America, research is needed to assess their psychometric properties among non-native English-speakers before they can with confidence be used in international research settings. No such research has yet been undertaken. The six studies reported in this paper, first, examine the psychometric performance of Saucier’s (1994) big-five Mini-Markers, in an English-using, multinational sample, then, having found its psychometric properties to be suboptimal, develop and validate an International English Mini-Markers.

The Mini-Markers, a short-form of Goldberg’s (1992) unipolar lexical big-five measure, is selected for assessment because it has proven to be one of the most psychometrically reliable (Mooradian & Nezlek, 1996) and frequently used brief big-five measures, being employed widely in personality research (Diefendorff and Richard, 2003, McCullough et al., 2002) and in applied settings across several disciplines, including health (Austin, Saklofske, & Egan, 2005), business (Erdheim, Wang, & Zickar 2006), and education (Moon & Illingworth, 2005). Comprising just 40 single-adjective personality descriptors originally selected for their psychometric qualities (Saucier, 1994), the Mini-Markers can be hypothesized to lend itself without translation to use with multinational samples that use English.