Guidelines for choosing between multi-item and single-item scales for construct measurement: a predictive validity perspective

Abstract

Establishing predictive validity of measures is a major concern in marketing research. This paper investigates the conditions favoring the use of single items versus multi-item scales in terms of predictive validity. A series of complementary studies reveals that the predictive validity of single items varies considerably across different (concrete) constructs and stimuli objects. In an attempt to explain the observed instability, a comprehensive simulation study is conducted aimed at identifying the influence of different factors on the predictive validity of single versus multiitem measures. These include the average inter-item correlations in the predictor and criterion constructs, the number of items measuring these constructs, as well as the correlation patterns of multiple and single items between the predictor and criterion constructs. The simulation results show that, under most conditions typically encountered in practical applications, multi-item scales clearly outperform single items in terms of predictive validity. Only under very specific conditions do single items perform equally well as multi-item scales. Therefore, the use of single-item measures in empirical research should be approached with caution, and the use of such measures should be limited to special circumstances.

Pricing to manage export channel relationships

“…We decided not to bundle them into a single variable so that we may assess the consequences of each of these pricing behaviors separately. In addition, it is important to note that the simple behaviors described by each item do not require multiple indicators to be assessed (for a discussion of this topic, see Diamantopoulos et al, 2012).…”

Ethnic inequalities in doctor-patient communication regarding personal care plans: the mediating effects of positive mental wellbeing

“…Together, these fourteen items generated high internal consistency (α = 0.91). Nevertheless, it was decided to treat the statements as single item measures, for several reasons: (a) weak effect sizes were expected, (b) the items were highly homogeneous (α > 0.90), (c) items were semantically redundant (due to high internal consistency), which can negatively affect multiple item measures, and (d) the population was diverse (Diamantopoulos et al 2012, Fuchs andDiamantopoulos 2009). Although multi item measures can outperform single items when the sample size is over a certain cut off (e.g., 6 > 50) (Diamantopoulos et al 2012), both approaches may perform equally well with very large samples (Cheung and Lucas 2014).…”

“…Fourthly, the absence of a direct effect of ethnicity on doctor interpretations don't require a direct association between X and Y, in order to demonstrate mediation (Hayes 2009(Hayes , 2013. Finally, the use of single item measures may generate concern (Diamantopoulos et al 2012), albeit evidence suggests they work similarly to multi item measures with very large samples (Cheung and Lucas 2014).…”

Do Brands "JUST DO IT"? Studying the Effectiveness of B2C Slogans

“…--brand consuming frequency: simple item ordinal categorization (very frequent consuming/user; moderate consuming/user; less consuming/user), using Diamantopoulos et al (2012) …”

“…--sympathy towards the brand: simple item ordinal categorization ("have sympathy towards the brand"; "the brand is indifferent"; "have antipathy towards the brand"), using Diamantopoulos et al (2012) indications. …”