Author(s) This pre-registration is currently anonymous to enable blind peer-review. It has 3 authors.
Pre-registered on 2022/11/15 01:39 (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? Robot knowledge test: Validation of a questionnaire to assess people's knowledge about robots.
Research question: How do knowledge, attitude, and anthropomorphism relate to people's willingness to help a robot?
Hypothesis:
(a) Influence of knowledge: Knowledge influences people's willingness to help, with high knowledge leading to participants being less willing to help.
(b) Influence of attitude: Positive attitudes lead to people being more willing to help, while negative attitudes lead to people being less willing to help.
(c) Influence of anthropomorphism: Participants' own opinion on how anthropomorphic robots are influences their willingness to help a robot in need, with high anthropomorphism scores leading to participants being more willing to help.
3) Describe the key dependent variable(s) specifying how they will be measured. Willingness to help: Participants read 8 different scenarios about a robot that needs help. Each scenario is presented two times in randomized order, accompanied by a picture of a robot (16 counterbalanced experimental trials in total). Participants decide how likely it would be that they help (5-point Likert scale).
4) How many and which conditions will participants be assigned to? The study has a 4 (domain: domain 1, domain 2, domain 3, domain 4) x 2 (robot: anthropomorphic, non-anthropomorphic) within-subjects design.
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. Robot Knowledge: Questionnaire (12 true, 13 false items). The robot knowledge scale was pre-registered under https://aspredicted.org/8HD_BZ5 and pre-tested.
Attitudes will be measured using the 3-factor measure, Robotic Social Attitude Scale (RoSAS) (Carpinella et al., 2017). Participants rate on a 9-point scale how closely they associate 18 different words with the category "robots".
Anthropomorphism will be measured as the participant's opinion on how anthropomorphic robots are. The Waytz instrument (Ruijten et al., 2019) will be used.
Willingness to help will be calculated as the mean across all experimental trials.
Hypothesis:
(a) Multiple linear regression models predicting willingness to help from knowledge while controlling for attitude, anthropomorphism score, gender (1= male; 2= female), age, and education (dummy-coded).
(b) Multiple linear regression models predicting willingness to help from the attitude score while controlling for knowledge, anthropomorphism, gender (1= male; 2= female), age, and education (dummy-coded).
(c) Multiple linear regression models predicting willingness to help from anthropomorphism while controlling for knowledge, attitude, gender (1= male; 2= female), age, and education (dummy-coded).
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. Inclusion criteria: fluent in German, at least 18 years old;
Exclusion criteria: participants who fail both attention checks will be excluded without compensation. Participants are informed of this at the beginning of the study.
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 are aiming to sample N = 500 valid cases, a representative sample (age and gender) for the German population.
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) 1. We statistically compare the helping scores for the two stimuli categories (anthropomorphic, and non-anthropomorphic).
2. Multiple linear regression models predicting willingness to help calculated separately for humanoid and machine-like items, from knowledge while controlling for attitude, anthropomorphism, gender (1= male; 2= female), age, and education (dummy-coded).
3. We investigate whether the helping scores, calculated as domain-specific, are statistically different.
4. Multiple linear regression models predicting willingness to help from knowledge while controlling for attitude, anthropomorphism score, gender (1= male; 2= female), age, and education (dummy-coded), and scenarios as a random intercept.
5. We exploratorily conduct Bayesian analysis for the main hypotheses.
6. We additionally collect data on participants' motivation to help/not help, their political orientation, as well as their risk and opportunity perception related to the use of robots for the domains presented.
