Effects of a Nonviolent Communication Program on Nursing Students.
SAGE Open, 12(3).
Kim, H. K., & Jo, H. K.
DOI : 10.1177/21582440221096139
Full article: link
Abstract:
The aim of this study was to examine the effect of NVC on improving the communication skills of nursing students. The study included a total of 117 participants, including a test group of 62 nursing students and a control group of 55 students. Eight sessions of data collection and programme adaptation were conducted. After the intervention, primary and secondary levels of anger were lower, while empathy and communication effectiveness were higher in the test group than in the control group.
Comment:
The authors use a pre-post experimental design, i.e. they test the students before and after the NVC programme. They also use a control group, which is essential to rule out test-retest bias. However, the control group is made up of students from a different university and town from the group receiving the NVC training. And although the authors state that the populations and socio-cultural contexts are similar, they do not measure this. It is therefore impossible to distinguish whether the differences found between the two groups are due to the NVC programme or to the fact that the students live and work in a different city.
What’s more, by conducting 6 independent statistical tests, one for each of their questionnaires (i.e. primary anger, secondary anger, interpersonal relations, empathy, self-esteem and communication effectiveness), the authors increase their chances of finding a difference between the two groups that is purely due to chance. In statistics, a simple method for correcting the significance threshold in multiple comparisons is the Bonferroni correction, so the p-value threshold they use at 0.05 could be divided by 6 to 0.008. According to their results table (Table 3), with this correction, only communication efficiency would still be significantly different between the two groups.
Furthermore, as the tests were explicit and the training issues fairly obvious, it seemed easy for the experimental group to fall into confirmation bias, by responding in such a way as to increase the scores for each test.
Perspective:
This study calls for close attention to be paid to statistics when making multiple comparisons and to the use of implicit tests to evaluate the effects of an educational programme.