Two types of tests used
- t-test was used for the normally distributed population. Most suited with large sample size and gives best results where the difference is the main issues. For example, states with the NPs being compared to those that do not have.
- Mann Whitney U test. This is best applied where the data is not normally distributed. It is tedious with large samples. It’s used when the maximum difference between two samples is expected.
- The use of Mann-Whitney tests indicates that, in the null hypothesis, there is no significant difference between the two populations where the samples were collected. The outcome of the data would mean that the two samples have similarities. In this case, the states with independent NPs were not different form those without the NPs in regard to access to care.
- The use of t-test to test normal distribution of the data revealed that there was no significant in the data set. The mean, mode and median of the data were almost equal meaning that the data was normally distributed. The Kurtosis test affirmed this by indicating that the distribution of healthcare access and status were relatively small.
- Hypothesis errors were present in the research. A large Mann- Whitney U indicates that the samples used were alike and hence the null hypothesis is not rejected. External validities in the research had an influence on the data collected and this affected the results and the hypotheses
- Type 1 error this would entail rejecting the null hypothesis while it’s true. This error would have made the researcher to conclude that there was a difference in states with NPs compared to those without NPs in healthcare provision. It would mean unequal healthcare provision. This would indicate a crisis in health provision and is much serious.
- Type 2 error. This entails accepting the alternative hypothesis while it should be rejected. This would have no serious effect. NPs distribution would not be affected in decision making.