Matching respondents based on age and sex

yuchinher — Mon, 22 Feb 2021 15:28:03 GMT

Hi,

I am working on a study in which I have to match the respondents based on their best friend's age and sex. An example looks like the following. I would like to use the respondent's age and sex and his/her best friend's age and sex to predict the number of hours a respondent goes to a party a week. However, to examine whether the effect is really from the best friend or from the sex and age of the best friend, I want to do random matching based on the best friend's age and sex. That is, I would like assign each respondents to a person who has the same age and sex as his/her best friend but is in fact not his/her best friend. For example, for the first respondent, he should be match with the best friend of the fourth respondent, and the second respondent should be match with the 8th, etc.

ID friend's sex friend's age sex age party hour/week

1 F 20 M 30 5h

2 M 21 F 29 0h

3 F 23 F 25 3h

4 F 20 M 24 1h

5 M 22 M 22 3h

6 F 20 F 21 5h

7 F 23 F 22 4h

8 M 21 F 20 5h

9 F 24 M 20 10h

10 F 24 F 27 8h

I would be really appreciated if anyone can help me regarding how to program this and to make a random matching using age and sex possible. Any advice or question are also welcome.

Best,

Re: Matching respondents based on age and sex

Reeza — Mon, 22 Feb 2021 16:55:55 GMT

PROC PSMATCH or search greedy match macro Mayo Clinic.

Re: Matching respondents based on age and sex

PGStats — Mon, 22 Feb 2021 18:42:25 GMT

Which of the following are true

1 - The friends are not part of the respondents

2- No two respondents have the same best friend

3- You want exact matches on sex and age

topic Re: Matching respondents based on age and sex in SAS Programming

Matching respondents based on age and sex

Re: Matching respondents based on age and sex

Re: Matching respondents based on age and sex