1. Case introduction
A certain unit wants to study whether a health product has an anti-fatigue effect on mice. Mice of the same species are paired according to their gender, age, and weight. There are 14 pairs in total, and two pups in each pair are Rats were randomly assigned to two different health food dosage groups, and the swimming time (min) of the mice was measured when the mice were loaded with 5% of their body weight, and whether there was a difference in the weight-bearing swimming time of the mice in different dosage groups.
2. Problem Analysis
If you want to study the difference between paired quantitative data, you can generally use the paired t test or the paired wilcoxon signed rank test, but the paired t test is generally suitable for data where the difference between the two sets of data is approximately subject to a normal distribution, and generally does not meet this premise , which can be analyzed using the paired Wilcoxon signed-rank test. After testing, the data in this case do not obey the normal distribution, so the paired Wilcoxon signed rank test was used for analysis.
3. Software operation and result interpretation
(1) Data import
1. Data format
First organize the data into the correct format, the number of paired data rows must be exactly equal and only two columns. The data format is as follows:
2. Import data
Upload the organized data to the SPSSAU system as follows:
The upload results are as follows:
(2) Paired wilcoxon signed rank analysis
- Software operation
The analysis path of using the paired wilcoxon signed rank test is to click [Experiment/Medical Research] → [Paired wilcoxon] and then analyze:
- Interpretation of results
As can be seen from the above table, the paired analysis of the mice in the middle dose group and the high dose group found that the statistic was 2.679, and the p value was less than 0.05, indicating that there was weight-bearing swimming between the mice in the middle dose group and the mice in the high dose group time is significantly different. And compared with the median, it can be seen that the median of the middle dose group (13.900) will be significantly higher than the median of the high dose group (9.250).
4. Conclusion
There is no difference in the weight-bearing swimming time of mice in different dose groups, because the data format is paired quantitative data, using paired t-test or paired wilcoxon signed rank test, but the difference between the tested paired data does not satisfy the normal distribution, Therefore, the paired wilcoxon signed rank test was used for research, and the model showed that there were significant differences in weight-bearing swimming time between different groups of mice. And the median (13.900) of the middle dose group will be significantly higher than the median (9.250) of the high dose group.
5. Knowledge Tips
1. For the paired data, the normality test is to detect the difference between the two columns of the paired data, not the normality test for the two columns of data.
2. Sometimes the results show that there are differences, but the median is the same. At this time, it is recommended to use the average value to represent the difference (it can be calculated by using subtotals). The possible reason for this situation is that the data has abnormal values;
3. Sometimes the results show that there is no difference, but the median is obviously inconsistent. At this time, it is recommended to use the average value to represent the difference (it can be calculated by subtotaling). The possible reason for such a situation is that the data has outliers.