Matching on one variable only
Sex Place all the IDs for Men together. There are 27 men. Randomly select half of the men to the control group and the other half to the experimental group. There will be 13 men in each group. One man will not be selected for either group.
Place all the IDs for Women together. There are 23 women. 11 women will be randomly selected to be in the control group, and 11 women will be selected to be in the experimental group. One women will not be placed in either group.
Age Place all the IDs for those 65+ together. There are 10 individuals in this age category. Randomly select 5 for the experimental and control groups.
Place all the IDs for those under 65 together. There are 40 individuals in this age category. Randomly select 20 for the experimental and 20 for the control group.
Location Place all the IDs for those who live in cities together. There are 28 individuals who live in cities. Randomly select 14 for the control group and 14 for the experimental group.
Place all the IDs for those who live in towns together. There are 11 individuals in this category. Randomly select 5 for the control group and 5 for the experimental group. One individual will not be selected for either group.
Place all the IDs for those who live in rural areas together. . There are 11 individuals in this category. Randomly select 5 for the control group and 5 for the experimental group. One individual will not be selected for either group.
Create matched samples for the variable location.
Use systematic sampling in each category to assign individuals to the control and experimental groups.
Matching on two variables
Age
and Sex
There will be 4 combinations of age and
sex:
Female Under
65
Male Under
65
Female 65+
Male 65+
Make
a list of all IDs for the first category (females under 65). There are 20 such
individuals. Randomly place ten individuals in the experimental and ten in the
control group.
Make
a list of all IDs for the second category (males under 65). There are 20 such
individuals. Randomly place ten individuals in the experimental and ten in the
control group.
Make
a list of all IDs for the third category (females 65+). There are 3 such
individuals. Randomly place one individual in the experimental and one in the
control group. One individual will not be selected for either group.
Make
a list of all IDs for the fourth category (males 65+). There are 7 such
individuals. Randomly place 3 individuals in the experimental and 3 in the
control group. One individual will not be selected for either group.
Sex
and Location
There will be 6 combinations of sex and
location:
Female City
Female Town
Female Rural
Male City
Male Town
Male Rural
Make
a list of all IDs for the first category (females, city). There are 15 such
individuals. Randomly place 7 individuals in the experimental and 7 in the
control group. One individual will not be selected for either group.
Make
a list of all IDs for the second category (females, town). There are 5 such
individuals. Randomly place 2 individuals in the experimental and 2 in the
control group. One individual will not be selected for either group.
Make
a list of all IDs for the third category (females, rural). There are 3 such
individuals. Randomly place one individual in the experimental and one in the
control group. One individual will not be selected for either group.
Do
the same for males
Males,
city 13
Males,
town 6
Males,
rural 8
In
total 4 individuals from our sample will not be matched.
Matching on three variables
Sex and Age and Location
There will be 12 combinations of sex, age and location:
|
|
Females |
Males |
||
|
|
Under 65 |
65+ |
Under 65 |
65+ |
|
City |
|
|
|
|
|
Town |
|
|
|
|
|
Rural |
|
|
|
|
|
|
Females |
Males |
||
|
|
Under 65 |
65+ |
Under 65 |
65+ |
|
City |
13 |
2 |
8 |
5 |
|
Town |
2 |
1 |
7 |
1 |
|
Rural |
5 |
0 |
5 |
1 |
Some of the categories have odd numbers, so not all of the individuals will be matched.
|
|
Females |
Males |
||
|
|
Under 65 |
65+ |
Under 65 |
65+ |
|
City |
12 |
2 |
8 |
4 |
|
Town |
2 |
0 |
6 |
0 |
|
Rural |
4 |
0 |
4 |
0 |
Eight individuals will have no match, and so will not be included in either the experimental or control group.
Create matched samples for all three variables.
Use systematic sampling in each category to assign individuals to the control and experimental groups.