Should we use single items? Better not☆
Section snippets
Thank you, but…
Keeping with Oscar Wilde's famous quote “there is only one thing in life worse than being talked about, and that is not being talked about”, we very much appreciate the comments on our paper “Selecting single items to measure doubly concrete constructs: A cautionary tale” (Sarstedt, Diamantopoulos, Salzberger, & Wilczynski, 2016) by Bergkvist (in press) and Rossiter (2016). Considering that some of the paper's results question their research findings on the appropriateness of single-item
Where have all the expert raters (and users) of C-OAR-SE gone?
Construct definition in C-OAR-SE ultimately depends on rational expert judgment; or as Bergkvist (in press) notes: “the role of expert judges should be to determine the nature of the construct (…)”. For example, to classify a construct as doubly concrete, expert raters (or a majority thereof) have to agree that the construct has a simple, clear object coupled with a single and single-meaning attribute (Bergkvist, L., 2016, Rossiter, J. R., 2002). If the expert raters agree on the doubly
To predict or not predict: And if not, how?
The notion that single items of doubly concrete constructs exhibit at least the same predictive validity as their multi-item counterparts is one of the cornerstones of Rossiter, J. R., 2002, Rossiter, J. R., 2011a, Rossiter, J. R., 2011b C-OAR-SE procedure and has been empirically tested by Bergkvist, L. and Rossiter, J. R., 2007, Bergkvist, L. and Rossiter, J. R., 2009. However, the relevance of the latter two studies for C-OAR-SE remains highly unclear. On the one hand, Bergkvist and Rossiter
Conclusion
We are thankful for Bergkvist's (in press) and Rossiter's (2016) comments but for reasons depicted above are hesitant to accept their criticisms. In fact, their rejoinder misses the central point of the debate. There are many appealing reasons to use single items (e.g., Fuchs & Diamantopoulos, 2009) but predictive validity is certainly not one of them. Regardless of whether a construct is a doubly concrete construct, a concrete construct, or even just a construct, single items lag behind
References (48)
- et al.
Why you must use my C-OAR-SE method, by John Rossiter: Review and commentary
Australasian Marketing Journal
(2015) - et al.
Understanding the influence of cues from other customers in the experience: A scale development and validation
Journal of Retailing
(2012) The C-OAR-SE procedure for scale development in marketing: A comment
International Journal of Research in Marketing
(2005)- et al.
Website performance and behavioral consequences: A formative measurement approach
Journal of Business Research
(2013) - et al.
Receptiveness of Black Americans to outdoor advertising
Journal of Business Research
(2011) - et al.
The use of partial least squares structural equation modeling in strategic management research: A review of past practices and recommendations for future applications
Long Range Planning
(2012) - et al.
Firm-employee-relationship strength — A conceptual model
Journal of Business Research
(2009) - et al.
Killing two birds with one stone: Cross-selling during service delivery
Journal of Business Research
(2014) The C-OAR-SE procedure for scale development in marketing
International Journal of Research in Marketing
(2002)Reminder: A horse is a horse
International Journal of Research in Marketing
(2005)
Why you must use my C-OAR-SE method
Australasian Marketing Journal
Consumer perception of taboo in ads
Journal of Business Research
Selecting single items to measure doubly concrete constructs: A cautionary tale
Journal of Business Research
Appropriate use of single-item measures is here to stay
Marketing Letters
The nature of doubly concrete constructs and how to identify them
Journal of Business Research
The predictive validity of multiple-item versus single-item measures of the same constructs
Journal of Marketing Research
Tailor-made single-item measures of doubly concrete constructs
International Journal of Advertising
Environmental uncertainty and forward integration in marketing: Theory and meta-analysis
European Journal of Marketing
A paradigm for developing better measures of marketing constructs
Journal of Marketing Research
Applied multiple regression/correlation analysis for the behavioral sciences
Guidelines for choosing between multi-item and single-item scales for construct measurement: A predictive validity perspective
Journal of the Academy of Marketing Science
Statistical analysis in psychology and education
The tension between strategy and execution: Challenges for international advertising research
Journal of Advertising Research
Publication bias in the social sciences: Unlocking the file drawer
Science
Cited by (37)
An off-target scale limits the utility of Short Warwick-Edinburgh Mental Well-Being Scale (SWEMWBS) as a measure of well-being in public health surveys
2022, Public HealthCitation Excerpt :However, measurement uncertainties reflect the lack of exact knowledge of the value of the measurand (e.g. person attribute values and item attribute values) and characterizes the dispersion of the values that reasonably could be attributed to it.30 Few items and/or mistargeted items imply low measurement precision.31 More items mean that we get more information about each respondent and especially if the number of well-targeted items increases the measurement uncertainties for estimates of person attribute values are reduced.
What explains the degree of internationalization of early-stage entrepreneurial firms? A multilevel study on the joint effects of entrepreneurial self-efficacy, opportunity-motivated entrepreneurship, and home-country institutions
2020, Journal of World BusinessCitation Excerpt :First, because we analyzed early-stage entrepreneurial firms’ internationalization, it was appropriate to focus on the impact of home-country institutions; however, host country institutions will play a greater role as these firms continue to grow internationally and this is why we call for future research to examine how ESE and opportunity-motivated entrepreneurship interact with both home- and host-country institutions as they continue to internationalize. Second, as some studies question the appropriateness of using a single item (Sarstedt, Diamantopoulos, & Salzberger, 2016), future research could use multidimensional measures of self-efficacy. Because we treat ESE as general rather than specific self-efficacy, it is appropriate to apply unidimensional and binary measures from the GEM survey.
Religiousness and environmental concern: A multilevel and multi-country analysis of the role of life satisfaction and indulgence
2018, Journal of Business ResearchCitation Excerpt :Nevertheless, potential criticisms arise from the fact that the WVS uses single-item scales for a substantial number of its measures. However, even critics of single-item measures acknowledge that there are situations in which the use of multi-item scales is not feasible due to budget restraints or questionnaire length (Sarstedt, Diamantopoulos, & Salzberger, 2016). Further, supporters of single-item scales demonstrate the validity of their measures in a substantial number of studies related to, e.g., trust (Johnson & Mislin, 2012) and life satisfaction (Diener et al., 2013).
Secondary agency conflicts: A synthesis and proposed measurement model
2018, Long Range PlanningCitation Excerpt :In this regard, we submit that the Shareholder Inequity index developed in this paper is more comprehensive based on its broader set of indicators that capture a relatively larger proportion of the secondary agency construct domain than prior approaches (see Table 1). Indeed, scale development research endorses the superiority of multi-item measures over single-item measures (Sarstedt, Diamantopoulos and Salzberger, 2016; Sarstedt, Diamantopoulos, Salzberger and Baumgartner, 2016; Sarstedt and Wilczynski, 2009) since by their nature they capture a more diverse scope of the construct's domain, thereby enabling a more accurate representation of the nature of the observed phenomenon (Hair et al., 2017); in the present context, potential secondary agency conflicts. Our new index facilitates the exploration of potential secondary agency conflicts in future studies and enables researchers to integrate individual findings more consistently and coherently into an overall framework.
- ☆
The authors would like to thank Florian Kolbe for his support with the analysis. Furthermore, the authors would like to thank Barry J. Babin and Mitch Griffin for their comments on an earlier version of this article.