'Nature & Personal Food Ratings' (AsPredicted #81808)
Author(s) Maria Langlois (INSEAD) - langlois@smu.edu Pierre Chandon (INSEAD) - pierre.chandon@insead.edu
Pre-registered on 2021/12/02 - 01:44 PM (PT)
1) Have any data been collected for this study already? No, no data have been collected for this study yet.
2) What's the main question being asked or hypothesis being tested in this study? We expect that exposure to nature will lead to healthier food consumption decisions when compared to exposure to urban environments. Additionally, we predict that nature increases the preference for healthy food, rather than alters the perceived healthiness of foods. Furthermore, we do not anticipate that the sequence of events, that is, whether individuals rate foods first vs. see the environmental (nature vs. urban) intervention first, will impact dietary choice outcomes.
3) Describe the key dependent variable(s) specifying how they will be measured. In a hypothetical lunchbox meal choice task, participants will choose a drink, main course, and side dish, which are framed as a free meal offering that they would be able to consume while enjoying the view depicted in one of the 2 scenes (i.e., their randomly assigned condition of nature or urban). Respondents will have the option to choose 1 out of 4 drinks (two of which are healthy & two of which are unhealthy), 1 out of 4 side dishes (two of which are healthy & two of which are unhealthy), and 1 out of 4 main courses (two of which are healthy & two of which are unhealthy).
4) How many and which conditions will participants be assigned to? 2x2 between-subjects design with Environmental Condition (Nature vs. Urban) as the main IV, and the order of tasks (rating foods first vs. environmental intervention first) as another IV
Participants will be randomly assigned to one of the two environmental conditions. The conditions will be displayed in the form of a photographic scene exhibiting either a nature view or an urban view
Participants will also be randomly assigned to either rate the healthiness of food/beverages (a set of 4 drinks, 4 main dishes, & 4 sides, for a total of 12 items) first and then see the environmental photo (intervention mentioned above) or to have the intervention first and then rate the food/beverages
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. We plan on conducting analyses which compare food choices across environmental contexts using binomial logistic regressions, with an interaction between environmental context (nature vs. urban) and each person's rating of the healthiness of the food. We will also look to see if the sequence of the independent variables, i.e., whether participants rated foods first vs. saw the environmental intervention first, impacts food choice outcomes. This means we'll run a binary logit with multiple interactions, including a 3-way interaction between environmental condition, food ratings, and the sequence of events. Furthermore, we will utilize conditional logit to account for the fact that the same individuals are making a 1 in 4 choice decision (e.g., one out of 4 drink options) on 3 occasions (for a drink, a main dish, and a side). We plan to conduct Hayes' PROCESS model 74 of moderated-mediation to investigate the pathway from environmental context to food choice, and to see what impact, if any, the environmental intervention has on food healthiness ratings as well as on the pathway of food healthiness ratings to food choice. Finally, as a robustness check, all these analyses will be done using the mean rating of each food's healthiness, as determined by a different sample of respondents (rather than in the respondent's own ratings of healthiness).
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. Participants will be excluded: if they indicate that they are not using an eligible device, if meta-data reveals an operating system or desktop resolution that is likely to be a cell phone/tablet (browser resolution height under 500), and if they failed the attention check (which consists of a recall question).
7) How many observations will be collected or what will determine sample size? No need to justify decision, but be precise about exactly how the number will be determined. The sample size was determined based on the effect size observed in our previous pre-registered online study, which utilized a control condition to address the alternative explanation of urban environments driving unhealthy eating (as opposed to natural environments driving healthier eating). Using the values from this previous study yielded n=1378. Given that we'll have 3 data points per person (a choice of drink, main dish, & side), we can divide that value by 3, for 460 participants. However, the planned interaction necessitates double the observations, yielding 920 participants. Therefore, we will recruit 920 participants online for this study.
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) We capture dieting behavior and sex for exploratory purposes.
