USEFUL STUFF - SAMPLE DATA DEVIATION TABLE
Sample deviation (confidence interval) depends on 3 factors:
- Sample size
- Confidence level
- Parameter value in sample
Confidence intervals decrease (sample deviation decreases) when:
- Sample size is increased
- Confidence level is decreased
- When the parameter value is close to 0% or 100% (confidence interval is biggest at 50%)
95% confidence level is the most widely accepted standard for evaluating study data deviation.
Sample size |
Confidence level 95% |
Parameter value, % |
||||||||||
5.0 |
10.0 |
20.0 |
30.0 |
40.0 |
50.0 |
60.0 |
70.0 |
80.0 |
90.0 |
95.0 |
||
50 |
Low |
0.8 |
3.3 |
10.0 |
17.9 |
26.4 |
35.5 |
45.2 |
55.4 |
66.3 |
78.2 |
84.8 |
High |
15.2 |
21.8 |
33.7 |
44.6 |
54.8 |
64.5 |
73.6 |
82.1 |
90.0 |
96.7 |
99.2 |
|
100 |
Low |
1.6 |
4.9 |
12.7 |
21.2 |
30.3 |
39.8 |
49.7 |
60.0 |
70.8 |
82.4 |
88.7 |
High |
11.3 |
17.6 |
29.2 |
40.0 |
50.3 |
60.2 |
69.7 |
78.8 |
87.3 |
95.1 |
98.4 |
|
200 |
Low |
2.4 |
6.2 |
14.7 |
23.7 |
33.2 |
42.9 |
52.9 |
63.1 |
73.8 |
85.0 |
91.0 |
High |
9.0 |
15.0 |
26.2 |
36.9 |
47.1 |
57.1 |
66.8 |
76.3 |
85.3 |
93.8 |
97.6 |
|
300 |
Low |
2.8 |
6.8 |
15.6 |
24.9 |
34.4 |
44.2 |
54.2 |
64.5 |
75.0 |
86.0 |
91.9 |
High |
8.1 |
14.0 |
25.0 |
35.5 |
45.8 |
55.8 |
65.6 |
75.1 |
84.4 |
93.2 |
97.2 |
|
500 |
Low |
3.3 |
7.5 |
16.6 |
26.0 |
35.7 |
45.5 |
55.6 |
65.8 |
76.2 |
87.0 |
92.7 |
High |
7.3 |
13.0 |
23.8 |
34.2 |
44.4 |
54.5 |
64.3 |
74.0 |
83.4 |
92.5 |
96.7 |
|
1000 |
Low |
3.7 |
8.2 |
17.6 |
27.2 |
36.9 |
46.9 |
56.9 |
67.1 |
77.4 |
88.0 |
93.5 |
High |
6.5 |
12.0 |
22.6 |
32.9 |
43.1 |
53.1 |
63.1 |
72.8 |
82.4 |
91.8 |
96.3 |
|
EXAMPLE:
It can be stated with 95% confidence that, at a sample size of 1000 and an observed value of 20%, the value of the parameter in the parent population will be in the interval of 17.6% to 22.6%.
To calculate sample error for parameter values not in the table, you can use the following calculator: