Dealing with missing data

[nonsi nkomo]

Hi

I am working on survey data with a total of 16 122 participants. I am interested on the effect of exercise on a persons BMI. The dependent variable, BMI was recorded for 7 099 participants and the independent variable, exercise was only recorded for 1 427. Other explanatory variables also have information for less than the total number of participants. I'm a bit weary about using functions like impute. Also I feel like there would be a loss in sample size power if i use listwise deletion. What would be the best way to go about removing these missing values?


Kind Regards
Nonsi Nkomo

[Clyde Schechter]

The handling of missing data is a complex topic, and many books and articles have been written on the subject. The best approach depends on both the process that generates the missing data and the type of analysis you wish to do with it. For an overview of the field, you could read Paul Allison's https://pdfs.semanticscholar.org/58d...c218e126e4.pdf.

[William Lisowski]

To add to Clyde's answer, there's a link to that document from https://statisticalhorizons.com/resources/articles with full citation information for the chapter.

Allison, Paul D. (2009) "Missing Data." Pp. 72-89 in The SAGE Handbook of Quantitative Methods in Psychology, edited by Roger E. Millsap and Alberto Maydeu-Olivares. Thousand Oaks, CA: Sage Publications Inc.

[nonsi nkomo]

Thank you all for the help.

Kind Regards
Nonsi Nkomo

[Carlo Lazzaro]

Nonsi:
see also: https://www.wiley.com/en-us/Survey+N...-9780471396277

[nonsi nkomo]

Thank you Carlo