Bert Weijters

Misresponse to Reversed and Negated Items in Surveys: A Review

There are important advantages to including reversed items in questionnaires (e.g., control of acquiescence, disruption of nonsubstantive responding, better coverage of the domain of content of a construct), but reversed items can also lead to measurement problems (e.g., low measure reliability, complex factor structures). The authors advocate the continued use of reversed items in measurement instruments but also argue that they should be used with caution. To help researchers improve their scale construction practices, the authors provide a comprehensive review of the literature on reversed and negated items and offer recommendations about their use in questionnaires. The theoretical discussion is supplemented with data on 1330 items from measurement scales that have appeared in Journal of Marketing Research and Journal of Consumer Research.

The Effect of Familiarity with the Response Category Labels on Item Response to Likert Scales

Surveys in the social sciences often employ rating scales anchored by response category labels such as “strongly (dis)agree” or “completely (dis)agree.” Although these labels may exert a systematic influence on responses since they are common to all items, academic research on the effect of different labels is surprisingly scarce. In order to help researchers choose appropriate category labels, we contrast the intensity hypothesis (which posits that response categories are endorsed less frequently if the labels are more extreme) with the familiarity hypothesis (which states that response categories are endorsed more frequently if the labels are more common in day-to-day language). In a series of studies we find consistent support for the familiarity hypothesis. Our results have important implications for the appropriate use of category labels in multilingual surveys, and we propose a procedure based on Internet search engine hits to equate labels in different languages in terms of familiarity.

Discriminant validity where there should be none. Positioning same-scale items in separated blocks of a questionnaire

In questionnaires, items can be presented in a grouped format (same-scale items are presented in the same block) or in a randomized format (items from one scale are mixed with items from other scales). Some researchers have advocated the grouped format because it enhances discriminant validity. The current study demonstrates that positioning items in separate blocks of a questionnaire may indeed lead to increased discriminant validity, but this can happen even in instances where discriminant validity should not be present. In particular, the authors show that splitting an established unidimensional scale into two arbitrary blocks of items separated by unrelated buffer items results in the emergence of two clearly identifiable but artificial factors that show discriminant validity.

Social Influences in Recruitment: When is Word‐of‐Mouth Most Effective?

We apply a policy‐capturing design to examine the conditions under which word‐of‐mouth is most effective in recruitment. The effect of monetary incentives is compared to other key characteristics of word‐of‐mouth (the source, recipient, and message content) that might affect its impact on organizational attractiveness. In a first study, unemployed job seekers (N = 100) were less attracted when they knew a monetary incentive was offered to the source of positive word‐of‐mouth. Conversely, they were more attracted when word‐of‐mouth was provided by a more experienced source (employee) and by a stronger tie (friend). These findings were replicated in a second study among employed job seekers (N = 213). These results offer various implications for how recruiting organizations might make effective use of word‐of‐mouth.

Development and test of an integrative model of job search behaviour

Research on job search and the theory of planned behaviour (TPB) has identified job search attitude, subjective norm, and job search self-efficacy as the most proximal determinants of job seekers’ search intentions and subsequently job search behaviours. However, we do not yet know how more distal individual differences (e.g., personality) and situational factors (e.g., social context) might help to predict these key TPB determinants of job search behaviour. In an integrative model of job search behaviour, we propose specific relationships between these distal variables and the TPB determinants, which in turn are expected to mediate the effects of individual differences and situational factors on job search behaviour. The hypothesized model is tested in a large representative sample of 1,177 unemployed Flemish job seekers using a two-wave design and provides a satisfactory fit to the data. Extraversion, conscientiousness, core self-evaluations, employment commitment, financial need, and social support are found to differentially relate to instrumental job search attitude, affective job search attitude, subjective norm, and job search self-efficacy. In addition, all distal variables are indirectly related to job search behaviour through their effects on the TPB variables. These results support our expanded and integrative model of job search behaviour.

Channels in the Mirror: An Alignable Model for Assessing Customer Satisfaction in Concurrent Channel Systems

Firms operating multiple channels as parallel routes to market face intense pressure to ensure superior customer satisfaction in their entire channel system. Relying on the structural alignment framework, the authors argue that to address this challenge, providers of concurrent channels should give priority to alignable channel attributes—attributes that have corresponding or “mirror” attributes in the other channels. These features are more salient to customers than nonalignable features and likely represent the origin of satisfaction evaluations in concurrent channel environments. Applying multigroup nested models using data from off-line and online shoppers, the authors empirically validate choice (assortment breadth and depth), charge (availability of fair prices), convenience (efficiency of the purchase process), confidence (security of transactions), and care (assurance of promised quality) as alignable channel facets. The resulting 5C model is superior to existing models in that it enables the unified capture of both off-line and online satisfaction, allowing a meaningful comparison across formats. Using alignable satisfaction facets enables managers to trace true differences in the satisfaction levels between channels. In particular, a channel’s share of investment should match its share of unexploited satisfaction potential. The 5C model also supports within-channel decisions by revealing the impact of the five facets on overall satisfaction with each format.

Methodological Issues in Cross-Cultural Research

Regardless of whether the research goal is to establish cultural universals or to identify and explain cross-cultural differences, researchers need measures that are comparable across different cultures when conducting cross-cultural studies. In this chapter, we describe two major strategies for enhancing cross-cultural comparability. First, we discuss a priori methods to ensure the comparability of data in cross-cultural surveys. In particular, we review findings on cross-cultural differences based on the psychology of survey response and provide suggestions on how to deal with these cultural differences in the survey design stage. Second, we discuss post hoc methods to ascertain data comparability and enable comparisons in the presence of threats to equivalence.

Measurement Models for Marketing Constructs

Researchers who seek to understand marketing phenomena frequently need to measure the phenomena studied. Yet, constructing reliable and valid measures of the conceptual entities of interest is a nontrivial task, and before substantive issues can be addressed, the adequacy of the available measures has to be ascertained. In this chapter, we discuss a wide variety of measurement models that researchers can use to evaluate the quality of their measures. It is assumed that, generally, multiple measures are necessary to capture a construct adequately. We first present the congeneric measurement model , in which continuous observed indicators are seen as reflections of an underlying latent variable , each observed variable loads on a single latent variable, and no correlations among the unique factors (measurement errors) are allowed. We contrast the congeneric measurement model with the formative measurement model , in which the observed measures cause the composite variable of interest, and we also consider measurement models that incorporate a mean structure (in addition to a covariance structure ) and extend the single-group model to multiple groups . Finally, we address three limitations of the congeneric measurement model (zero loadings of observed measures on non-target constructs, no correlations among the non-substantive components of observed measures, and the assumption of continuous, normally distributed indicators) and present models that relax these limiting assumptions.

Analyzing Policy Capturing Data Using Structural Equation Modeling for Within-Subject Experiments (SEMWISE):

We present the SEMWISE (structural equation modeling for within-subject experiments) approach for analyzing policy capturing data. Policy capturing entails estimating the weights (or utilities) of experimentally manipulated attributes in predicting a response variable of interest (e.g., the effect of experimentally manipulated market-technology combination characteristics on perceived entrepreneurial opportunity). In the SEMWISE approach, a factor model is specified in which latent weight factors capture individually varying effects of experimentally manipulated attributes on the response variable. We describe the core SEMWISE model and propose several extensions (how to incorporate nonbinary attributes and interactions, model multiple indicators of the response variable, relate the latent weight factors to antecedents and/or consequences, and simultaneously investigate several populations of respondents). The primary advantage of the SEMWISE approach is that it facilitates the integration of individually varying policy capturing weights into a broader nomological network while accounting for measurement error. We illustrate the approach with two empirical examples, compare and contrast the SEMWISE approach with multilevel modeling (MLM), discuss how researchers can choose between SEMWISE and MLM, and provide implementation guidelines.

To Be Yourself or to Be Your Ideal Self?: Outcomes of Potential Applicants’ Actual and Ideal Self-Congruity Perceptions

Recruitment research on person–organization fit has typically focused on organizations’ fit with potential applicants’ actual self, not considering other possible self-images. Based on image congruity theory, we investigate how actual and ideal self-congruity relate to application intentions and intentions to spread word-of-mouth. In a first study, conducted in Belgium, actual and ideal self-congruity related positively to both outcomes. The relation with application intentions was equally positive for actual and ideal self-congruity. Ideal self-congruity showed a stronger positive relation with word-of-mouth intentions. A second study replicated these findings in the United States and tested for social adjustment concern (need to impress others) as a moderator. As social adjustment concern increased, relations of both outcomes with ideal (actual) self-congruity were stronger (weaker).