#72308 | AsPredicted

As Predicted:Opinions on military spending: effects of info provision & answer options (#72308)


Created:       08/06/2021 07:32 PM (PT)

This is an anonymized version of the pre-registration.  It was created by the author(s) to use during peer-review.
A non-anonymized version (containing author names) should be made available by the authors when the work it supports is made public.

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?
The main questions being asked here are as follows.

(i) Are people less likely to say the U.S. is spending too little on military and defense if presented with information emphasizing the negative impacts of this spending, compared to if they are presented with a list of justifications for the spending?

(ii) Are people less likely to say the U.S. is spending too little on military and defense if presented with information emphasizing the (large) magnitude of current military spending, compared to if they are presented with information that makes the spending sound smaller in magnitude (by changing how the spending is presented/contextualized)?

(iii) Does the availability of only two answer options artificially polarize responses, compared to if a third, neutral answer option is available to choose?

3) Describe the key dependent variable(s) specifying how they will be measured.
The key dependent variable will be the answer people give to the main question of the experiment.*

For half the participants, the main question will be: Do you think we are spending too little or too much on national defense and military?

For the other half of the participants, the main question will be: Do you think we are spending too little, about the right amount, or too much on national defense and military?

*In section 5) we describe how we code this variable numerically.

4) How many and which conditions will participants be assigned to?
Each participant will be randomized to one of 16 conditions (2 sets of answer options x 8 different information presentations). The 8 different information presentations have names and content as follows:
(i) noInformation: no information will be given
(ii) withDollarAmountOnly: current military spending will be given as a dollar amount only
(iii) percentOfGDP: current military spending will be given as a percentage of GDP
(iv) percentOfAllUSSpending: current military spending will be given as a percentage of all U.S. spending
(v) ofAllWorldSpending: current military spending will be given as a percentage of military spending around the world
(vi) percentOfDiscretionaryUSSpending: current military spending will be given as a percentage of *discretionary* U.S. spending
(vii) loomingThreats: a list of some prominent threats (justifying military spending) will be given instead of information on current spending
(viii) disastrousWars: a list of some prominent wars will be given along with the argument that these wars were futile

The exact wording of these conditions can be seen in the experiment code here: https://www.guidedtrack.com/programs/17356/edit

5) Specify exactly which analyses you will conduct to examine the main question/hypothesis.
The main research questions were listed in section 2). Below, we outline the approaches we'll take to exploring each one in turn. When we mention the "score," this refers to the mean answer to the main research question (where "too much" will be coded as -1, "about right" [where applicable] will be coded as 0, and "too little" will be coded as 1).

To address question (i), we will compare the mean score in the loomingThreats condition to the mean score in the disastrousWars condition using a t-test. We expect the mean score to be lower in the latter condition as it emphasizes the negative outcomes of that spending.

To address question (ii), we will compare the mean score in the percentOfGDP condition to the mean score in the percentOfDiscretionaryUSSpending condition using a t-test. We expect the mean score to be lower in the latter condition as it emphasizes the already large magnitude of the spending (as a proportion of discretionary U.S. spending). Additionally, we will pool both the conditions that make the magnitude of military spending sound large (i.e., the ofAllWorldSpending and percentOfDiscretionaryUSSpending conditions) and use a t-test to compare the mean score across those conditions to the mean score across the conditions that do not make the magnitude of military spending sound as large (i.e., the percentOfGDP and percentOfAllUSSpending conditions), as well as comparing both these conditions to the noInformation and withDollarAmountOnly conditions.

To address question (iii), we will use a t-test to compare the proportion of people saying "too little" in the conditions where there were two answer options, versus the conditions where there were three answer options. We expect that a higher proportion will say "too little" in the conditions where only two options are provided.

6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations.
Due to the structure of the questions, no respondents will be labeled as outliers.

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 stop data collection after at least 1344 participants have completed the experiment. (As long as the dropout rate does not vary significantly across conditions, this should result in 84 participants per condition. However, for pragmatic reasons, the criterion for stopping the experiment will be after collecting at least 1344 participants, rather than collecting 84 per condition.)

8) Anything else you would like to pre-register?
(e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?)

As a secondary analysis, we will run a two-way analysis of variance (ANOVA) to assess the effects of presented information and answer format on the question score.

We may also conduct exploratory analyses.