Random distribution is one way of achieving even distribution in
surveys. Use random distribution to randomly show a different question or
concept to different survey participants.
The randomization functionality in survey groups is sufficiently random
to accomplish what you're looking for from even distribution, especially with a
large enough sample size. Alida has tested random distribution in survey groups
using different sample sizes. This topic summarizes the test results and gives
you an idea of what to expect with your own distributions.
Test setup
- The survey group contained
5 questions.
- The survey group was set
to display the questions in random order.
- The number of objects to display was set to 1.
- The survey was tested with
the following sample sizes:
- 50 responses
- 100 responses
- 1,000 responses
- 10,000 responses
- For each sample size, the
test was run 5 times.
Results summary
In a perfectly even distribution, each question would be displayed to
participants 20% of the time. Testing revealed that random distribution
achieved a split very close to the 20% ideal.
50 responses
| Sample Size=50
|
Question 1
|
Question 2
|
Question 3
|
Question 4
|
Question 5
|
Total
|
| Test run 1
|
8
|
10
|
9
|
12
|
11
|
50
|
| Test run 2
|
7
|
9
|
7
|
14
|
13
|
50
|
| Test run 3
|
13
|
7
|
13
|
10
|
7
|
50
|
| Test run 4
|
4
|
15
|
10
|
11
|
10
|
50
|
| Test run 5
|
5
|
12
|
10
|
10
|
13
|
50
|
| Sum of all runs
|
37
|
53
|
49
|
57
|
54
|
250
|
| % times question was displayed to
participants
|
14.80%
|
21.20%
|
19.60%
|
22.80%
|
21.60%
|
100.00%
|
100 responses
| Sample Size=100
|
Question 1
|
Question 2
|
Question 3
|
Question 4
|
Question 5
|
Total
|
| Test run 1
|
17
|
23
|
18
|
25
|
17
|
100
|
| Test run 2
|
22
|
24
|
21
|
22
|
11
|
100
|
| Test run 3
|
23
|
17
|
18
|
19
|
23
|
100
|
| Test run 4
|
14
|
20
|
19
|
23
|
24
|
100
|
| Test run 5
|
25
|
21
|
16
|
17
|
21
|
100
|
| Sum of all runs
|
101
|
105
|
92
|
106
|
96
|
500
|
| % times question was displayed to
participants
|
20.20%
|
21.00%
|
18.40%
|
21.20%
|
19.20%
|
100.00%
|
1,000 responses
| Sample Size=1000
|
Question 1
|
Question 2
|
Question 3
|
Question 4
|
Question 5
|
Total
|
| Test run 1
|
210
|
202
|
202
|
216
|
170
|
1,000
|
| Test run 2
|
199
|
185
|
208
|
218
|
190
|
1,000
|
| Test run 3
|
209
|
186
|
190
|
204
|
211
|
1,000
|
| Test run 4
|
210
|
195
|
213
|
190
|
192
|
1,000
|
| Test run 5
|
188
|
190
|
214
|
207
|
201
|
1,000
|
| Sum of all runs
|
1,016
|
958
|
1,027
|
1,035
|
964
|
5,000
|
| % times question was displayed to
participants
|
20.32%
|
19.16%
|
20.54%
|
20.70%
|
19.28%
|
100.00%
|
10,000 responses
| Sample Size=10,000
|
Question 1
|
Question 2
|
Question 3
|
Question 4
|
Question 5
|
Total
|
| Test run 1
|
2,003
|
2,093
|
1,964
|
1,937
|
2,003
|
10,000
|
| Test run 2
|
1,960
|
1,961
|
1,961
|
2,056
|
2,062
|
10,000
|
| Test run 3
|
1,963
|
1,998
|
1,996
|
2,062
|
1,981
|
10,000
|
| Test run 4
|
1,947
|
1,978
|
2,052
|
2,038
|
1,985
|
10,000
|
| Test run 5
|
2,069
|
1,928
|
1,954
|
1,977
|
2,072
|
10,000
|
| Sum of all runs
|
9,942
|
9,958
|
9,927
|
10,070
|
10,103
|
50,000
|
| % times question was displayed to
participants
|
19.88%
|
19.92%
|
19.85%
|
20.14%
|
20.21%
|
100.00%
|