'The Prevalence and Correlates of Residential School Denialism in Canada'
(AsPredicted #148,781)
Download PDF

---

---

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 hypothesize that providing factual information about the history of residential schools in Canada will reduce the likelihood of (a) participants affirming statements made by denialists (those who deny the reality or legacy of Canada's Indian Residential School System) and (b) responding "I haven't thought too much about this/ Don't know" when evaluating those statements.

3) Describe the key dependent variable(s) specifying how they will be measured.
Our outcome measures are based on respondents' expressed agreement with nine items related to denialist claims. We will first conduct reliability and dimensionality analyses to determine whether the items scale together well. We will determine which items to include in our main outcome scale based on this analysis. If we find that the items contain sub-scales, we will treat the sub-scales as distinct outcomes and analyze them separately. (For robustness, we will report analyses using all nine items in a scale as an outcome as well, at least in the Supplementary Materials (SM)).

This first outcome will only be based on respondents' non-missing responses. To determine if treated respondents are less likely to respond "I haven't thought too much about this/ Don't know," we will calculate a second outcome variable as the proportion of scale items indicating "I haven't thought too much about this/ Don't know."

4) How many and which conditions will participants be assigned to?
Participants are assigned with equal probability to one of two conditions:

Condition 1: Participants read a short informational text describing the history of residential schools prior to indicating their agreement/disagreement with statements related to denialism.

Condition 2: Participants do not read any text prior to indicating their agreement/disagreement with statements related to denialism.

5) Specify exactly which analyses you will conduct to examine the main question/hypothesis.
We will use difference of means/ OLS to examine the effect of treatment on our main outcome measure (and either report our second analysis of "I haven't thought too much about this/ Don't know" responses in the body of the paper or the SM). If the experiment is balanced, we will report the Average Treatment Effect (ATE) using a simple difference of means test and present this either in the body of the paper or the SM. We will also use OLS regression with controls to address any imbalance between the treatment and control groups as well as to reduce noise and increase the precision of our estimates. Further, to understand the correlates of Residential School denialism we will include the following socio-demographic variables that are standard in social science research: age, gender, language, region, race/ethnicity, and party identification, and education. We have also included the following variables to conduct exploratory analyses on the relationship between denialism and pre-treatment socio-demographic and political attitudes/characteristics: immigration status, income, vote choice, historical/political knowledge, religion/ religiosity, ideological self-placement, political interest, trust in media, prejudice (based on feeling thermometer scores), anti-Indigenous resentment, in-group identification, and benevolent racial attitudes. We may not necessarily present all the variables in the body of the paper in a single model as this is likely an overspecified, "kitchen sink" model. We may also conduct exploratory moderation analyses by interacting covariates with the treatment indicator. We will use theory and standard diagnostic tools (VIF test, tests of model fit) to determine which items to present in the body of the paper. All variables will be made available for other scholars seeking to conduct exploratory analyses/replication efforts.

For robustness, we may examine individual items as outcomes and/or use multinomial logistic regression to compare respondents' likelihood of responding agree vs. disagree vs. "don't know" to each item.

6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations.
Outliers will not be excluded from the analysis. We may conduct exploratory sub-group 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.
Our sample size is determined by our budget, which is $4,800 CAD. The cost per complete depends on how long the survey takes to complete and the quality of the data. We expect to collect between 1,700 to 2,500 completes.

8) Anything else you would like to pre-register?
(e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?)
We will flag respondents as low-quality if they (a) complete the survey in less than 1/3 the median time (measured separately in each treatment arm); (b) fail a pre-treatment attention check; or (c) are identified as bots by Qualtrics. We will conduct the analyses described above on both the full sample and the sample of only high-quality responses and report results using each sample either in the body of the paper or in the SM.