Chapter 5: Analysing the Data Part II : Inferential Statistics

# Step 5: Interpretation of Your Result.

The main task within this section is to now take your decision to retain or reject Ho and apply it back to the original research question. There are two tasks involved. First, you should be able to make a summary statement that interprets the results of the statistics in terms of what was analysed. In essence you are specifically answering the research hypotheses that you stipulated.

For example, (these are taken from published reports):

"The mean percent reduction of SS at 24 and 48 h was statistically significant in comparison to the normal decrease observed during the pre-drug SS period (t(13) = 7.47, p < 0.001)."

"There was no relationship between BP and SD scores (r (98) = .14, ns), or between SD and LC scores (r (98) = -0.10, ns). However, a significant correlation between BP and LC scores was found (r (98) = -0.47, p < 0.001)."

"The two groups (N = 1921) were homogeneous according to sex ( 2 (1) = 1.81, p > 0.05), but not according to age ( 2(1) = 14.24, p < 0.001)."

"There was, in general, little relationship between population size and pace. On the MANOVA, the WilkÕs lambda for population size indicated a non-significant trend [approximate F(8,42) = 1.90, p < 0.085]."

The second part of Step 5 is to consider the broader implications of your findings. Take your decision about whether you have a real result or not back to the research question or hypothesis that you were initially interested in. Speculate as to the implications of this wonderful finding in terms of the theory that formed the basis of the research in the first place. In this course, we would like to encourage you to emphasise, or look for, psychological implications. Sometimes findings have important implications for some theory, or perhaps some practical implications such as a new cure or therapy! Perhaps there are longer term implications? Perhaps the finding, even though 'significant', has little relevance for anything? (Often seems to be the case!)