Are taller people less charitable?

Methods

Summary

This research will primarily be survey-based, and thus the primary statistical tools for data analysis would be multiple regression techniques, spotlight analyses, and mediation techniques. (See below for specific details.) Generally, I will aim to recruit participants that are relatively representative of their respective country, I will try to minimize bias and keep as many things constant as possible, and I will aim to rule out alternative explanations. I predict that physical height should reduce charitable giving through feelings of dominance -- but it may reduce giving through some other means! Using proper statistical procedures and controlling for other outcomes would help me statistically lead to one conclusion or another.

Challenges

Statistically, it would be impossible to determine cause and effect. That is, I simply would not be able to say that physical height causes or impacts charitable giving -- it could very well be your height could be an effect of charity. (Donate more, and you become taller!) Of course, this is reasonably unlikely, but statistically, I would be unable to claim either way, and the data would only suggest -- not conclusively -- that physical height impacts charitable giving.

Another challenge would be the Chinese sample. My plan is to translate the Chinese questionnaire into Chinese, but this would obviously create some biases when comparing across samples from different countries (usually, researchers should keep as many things the same as possible). But given that English language proficiency in China is generally low (or lower relatively), this would be unavoidable.

Finally, because I will be conducting online surveys, this would exclude individuals who are not online. Perhaps people who are taller/shorter or donate more/less simply don't use the Internet. However, I will use statistical procedures and weighting to ensure that the samples in each country are relatively consistent with the overall demographics in the population. For example, if a country is 60% men and 40% women, I will aim to recruit 60% male participants and 40% female participants in that country.

Pre Analysis Plan

I will analyze using multiple regression techniques in order to determine the relationships between physical height, Social Dominance Orientation, and charitable giving. I will also statistically control for other possible variables such as age and income in order to strengthen the view that it is people's heights that impact their charity through feelings of dominance, and not through another means.

Mediation techniques will primarily follow Preacher and Hayes' (2008) bootstrapping protocols in order to properly draw on the sample and reduce bias. It is expected that people's physical height should reduce their charitable giving, and this effect should be mediated by greater feelings of dominance, as indicated by higher scores on Social Dominance Orientation.