'Effects of Framing on Anti-Fat Bias and the Role of Pragmatic Reasoning' (AsPredicted #86578)
Author(s) This pre-registration is currently anonymous to enable blind peer-review. It has 2 authors.
Pre-registered on 2022/01/30 - 05:37 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? Previous work suggests that framing fatness in positive or negative ways affects the perceived health risk of being fat, opinions on public policy related to fatness, and anti-fat attitudes (Frederick et al., 2016a, 2016b, 2020; Saguy et al., 2014). In a norming study, we adapted news article stimuli from previous studies to specifically vary whether fatness is framed as a disease and/or equal-rights issue. We also developed a novel anti-fat bias scale. The norming study mapped the entailments of each frame and validated the anti-fat bias scale.
In the present study, we are investigating whether the frames from the norming study affect people's anti-fat bias, and if so, whether these framing effects are driven by pragmatic reasoning about the implications of the framing language. Participants will read a fictionalized news article that frames fatness as a disease or not and as an equal-rights issue (fat rights) or not. The order of the disease and fat-rights content will be counterbalanced. After reading the article, participants will complete the anti-fat bias scale, indicate whether or not they thought the article influenced their responses, and if so, which part.
We expect that the strongest effect on anti-fat bias will be elicited by the 'congruent' frames (i.e., anti-fat bias will be highest for the pro-disease/anti-fat-rights frame and lowest for the anti-disease/pro-fat-rights frame). We expect anti-fat bias to be intermediate for the 'incongruent' frames (pro-disease/pro-fat-rights, anti-disease/anti-fat-rights). We will also explore whether the magnitude of these framing effects differs by the type of anti-fat bias. Our norming study identified three types, captured by different subsets of anti-fat bias questions: emotion-based attitudes, human values, and health-related concerns. Finally, we will explore whether the magnitude of the framing effects is moderated by explicit recognition of the influence of the frames. Following previous research on linguistic framing (Flusberg et al., in press), we expect that the framing effects will be stronger in participants who cite the framing language as influential than in those who do not.
3) Describe the key dependent variable(s) specifying how they will be measured. ANTI-FAT BIAS [24 statements; Likert scale rating from 1 = strongly disagree to 7 = strongly agree] e.g., I would find it highly objectionable to see a fat person being teased or mistreated
PRAGMATIC REASONING [1 question, yes or no] Do you think the New York Times article you read had any influence, even slightly, on the personal beliefs about fat people that you just expressed? [If yes:] Which part of the New York Times article most influenced your personal beliefs about fat people? Copy that part from the article below and paste it into this text box:
DEMOGRAPHIC VARIABLES [7 questions; age (continuous); gender (multiple choice); race (multiple choice, select all that apply); political ideology (2 questions: multiple choice; Likert scale rating from -50 = very liberal to 50 = very conservative); highest level of education (multiple choice); annual household income (multiple choice)]
BODY WEIGHT SELF-IDENTIFICATION [3 questions; Do you consider yourself fat, obese, or overweight? (select all that apply); If you chose more than one descriptor, which do you most identify with? (multiple choice); Have you ever experienced discrimination because of your weight? (yes or no)]
4) How many and which conditions will participants be assigned to? Participants will be assigned to one of two disease frame conditions (pro/anti) and one of two fat-rights frame conditions (pro/anti). The order in which the disease and fat-rights content will be presented in the fictionalized news article will be counterbalanced.
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. We will first assess the reliability of the anti-fat bias scale and the three factors identified in the norming study (emotion-based attitudes, human values, health-related concerns). If the reliability of the factors is acceptable (Cronbach's alpha > .7), we will then calculate each participant's mean score for each factor.
The initial analysis will be a 2 (disease frame) x 2 (fat-rights frame) x 2 (order of frames) ANOVA on anti-fat bias scores. If there are no order effects, we will collapse across the two orders and conduct a 2 (disease) x 2 (fat-rights) x 3 (bias type) mixed ANOVA. If bias type interacts significantly with either of the other factors, we will conduct separate 2 (disease) x 2 (fat-rights) ANOVAs for each bias type.
To test whether any framing effects are moderated by explicit recognition of the influence of the framing language, we will first code whether participants cited the disease framing language as most influential, the fat-rights framing language as most influential, or neither as influential in their responses. Then we will conduct a 2 (disease) x 2 (fat-rights) x 3 (cited language) ANOVA on anti-fat bias scores, separately for each bias type for which significant framing effects were observed in the main analysis.
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. Two attention checks will be presented. At the beginning of the experiment, participants will be instructed to "check the option 'Other' below and enter the number 8 in the text box of this option." In the demographics section, participants will be asked to select "disagree" from a 5-point Likert scale. Participants who fail the first attention check will be prevented from completing the experiment, and data from participants who fail the second attention check will be excluded from analyses.
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. We will seek 400 participants pre-exclusion (100 per condition). Participants will be recruited through Amazon Mechanical Turk (MTurk), using the CloudResearch participant-sourcing platform. The sample will be limited to MTurk users in the United States who have an approval rating of at least 95% on 100 prior studies. Participants in our norming study will be ineligible.
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) Body weight self-identification and demographic information will be collected for exploratory purposes.
