A discussion on the results of Mcnemar's test by SPSS statistics

Objective To compare the results of Mcnemar's test by hand calculation and SPSS statistics, and to explain reasons for the difference. Methods Application examples of paired design data in related books were selected, and two calculation methods were used: the first was a hand calculation using Mcnemar's test formula, and the second was a calculation by SPSS. Difference of the results done by the two methods were compared. Results The result by hand calculation was P=0.001, while the result by SPSS was P=0.002. The results by the two methods were different. Conclusion SPSS statistics adopts an exact test in the calculation of 2 × 2 contingency tables with paired design data, in which a principle of bionominal distribution is applied. Despite the fact that the result differs from that of hand calculation, neither method will affect the final statistical inference.