David Gal

A Mouth-Watering Prospect: Salivation to Material Reward

The term “hunger” and terms referring to its physiological correlates, notably “salivation,” are used to refer to desire for material rewards across languages and cultures. Is such usage is “merely metaphorical,” or can exposure to material reward cues evoke a salivary response? Results of an experiment show that individuals salivate to money when induced to experience a low power state but not when induced to experience a high power state. A second experiment shows that men salivate to sports cars when primed with a mating goal but not in a control condition. These findings suggest that salivary secretion is stimulated by material rewards in the presence of a highly active goal to obtain the rewards and that the motivation to acquire material rewards might more closely resemble physiological hunger than previously assumed. Implications for material addictions and decision making and directions for future research are discussed.

Real Men Don't Eat Quiche: Regulation of Gender-Expressive Choices by Men

Everyday items are imbued with subtle yet pervasive gender associations. For instance, sour dairy products and products with rounded edges tend to be perceived as relatively feminine, whereas meat and products with sharp edges tend to be perceived as relatively masculine. In a series of studies, we find that men are more likely to choose gender-congruent options (masculine foods and angular-shaped items) when they have unconstrained time and attentional resources than when these resources are constrained. In contrast, women’s choices tend to not be affected by time or attentional resource availability. Our findings suggest that men experience a conflict between their relatively intrinsic preferences and gender norms and that they tend to forgo their intrinsic preferences to conform to a masculine gender identity (when they have sufficient resources to incorporate gender norm information in their choices). Women, on the other hand, appear to be less concerned with making gender-congruent choices.

Answering the Unasked Question: Response Substitution in Consumer Surveys

Researchers and practitioners alike frequently survey consumers to gain insights into their attitudes, preferences, and beliefs. The authors propose a potentially pervasive, but as of yet unidentified, source of bias in survey responding. Specifically, they propose that respondents’ answers to questions might sometimes reflect attitudes that respondents want to convey but that the researcher has not asked about, a phenomenon termed “response substitution.” The authors examine this proposition in a series of three experiments that demonstrate the phenomenon, provide support for the process account, and identify boundary conditions. They also discuss general theoretical, methodological, and practical implications as well as specific implications for research on attitudes and contingent valuation.

Blinding Us to the Obvious? The Effect of Statistical Training on the Evaluation of Evidence

Statistical training helps individuals analyze and interpret data. However, the emphasis placed on null hypothesis significance testing in academic training and reporting may lead researchers to interpret evidence dichotomously rather than continuously. Consequently, researchers may either disregard evidence that fails to attain statistical significance or undervalue it relative to evidence that attains statistical significance. Surveys of researchers across a wide variety of fields (including medicine, epidemiology, cognitive science, psychology, business, and economics) show that a substantial majority does indeed do so. This phenomenon is manifest both in researchers’ interpretations of descriptions of evidence and in their likelihood judgments. Dichotomization of evidence is reduced though still present when researchers are asked to make decisions based on the evidence, particularly when the decision outcome is personally consequential. Recommendations are offered. This paper was accepted by Yuval Rot...

Statistical Significance and the Dichotomization of Evidence

ABSTRACT In light of recent concerns about reproducibility and replicability, the ASA issued a Statement on Statistical Significance and p-values aimed at those who are not primarily statisticians. While the ASA Statement notes that statistical significance and p-values are “commonly misused and misinterpreted,” it does not discuss and document broader implications of these errors for the interpretation of evidence. In this article, we review research on how applied researchers who are not primarily statisticians misuse and misinterpret p-values in practice and how this can lead to errors in the interpretation of evidence. We also present new data showing, perhaps surprisingly, that researchers who are primarily statisticians are also prone to misuse and misinterpret p-values thus resulting in similar errors. In particular, we show that statisticians tend to interpret evidence dichotomously based on whether or not a p-value crosses the conventional 0.05 threshold for statistical significance. We discuss implications and offer recommendations.

When Sex and Romance Conflict: The Effect of Sexual Imagery in Advertising on Preference for Romantically Linked Products and Services

Sex is ubiquitous in advertising, yet little research has explored the effect of exposure to sexual imagery on preferences. Although sex and romance tend to go together in real-world relationships, the authors find that exposure to sex-based ads decreases preference for romantically linked products and services in men. Furthermore, the authors find that the effect is one directional, such that exposure to romantic imagery in ads does not decrease mens preference for sex-related products. Finally, the authors find that exposure to sex-based ads does not lead to a decreased preference for romantically linked products in women. The authors explain this pattern of results through the relatively opportunistic nature of the sex drive in men. They close with a discussion about implications for theories of fundamental motives and for the effect of sex-based advertisements on dating and relationships.

Let Hunger Be Your Guide? Being Hungry before a Meal Is Associated with Healthier Levels of Postmeal Blood Glucose

Accumulating evidence suggests that when people eat might be as important for their health as what or how much they eat. Consumers initiate eating behavior in response to a number of factors including time of day, food availability, and hunger. The present research examines whether feeling hungry might reflect a time at which it is relatively healthy for individuals to initiate eating. Specifically, the present research examines whether differences in hunger predict differences in an important health measure, namely, postmeal blood glucose. The results show that post-meal blood glucose is highest when people are not at all hungry and relatively lower when people are moderately hungry or very hungry. When people are very hungry, postmeal glucose is not lower and possibly slightly higher than when people are moderately hungry. The findings are consistent with the hypothesis that, from a health standpoint, it is beneficial to eat when moderately hungry.

Rejoinder: Statistical Significance and the Dichotomization of Evidence

including frequentist and NHST-based inference. As has been intensively discussed elsewhere, we are likely to be increasingly working with extensive volumes of fine-scale data on the systems we study. It has also been noted that “big data needs big models” (Gelman 2014). These big models, including models derived frommachine learning methods, as well as flexible procedures deriving from classical statistics such as semiparametric, empirical likelihood, dimension reduction, and localized methods, can be powerful tools for improving the properties of NHST. Recent work on high dimensional inference is providing new tools to build such models while not saturating the models to the point where parameter estimates become meaningless. However, most applied researchers and many statisticians are not using these new tools to their full potential. The findings ofMcShane andGalmake clear that in terms of communication, training, andmethods development, there is still a lot of room to grow.