How to Perform t-tests & Analyze Results

Starting with a Statement of the Hypotheses Being Tested

The report should start with a statement of the hypotheses being tested, and then structured into two parts: results and discussion. In the results start with a description of your observations and the figures before the figures. The disucssion is where you make meaning of the findings, link to other studies, state whether you accept or reject the hypothesis and look forward to future applications or studies The t-test is used to determine if the the two means are statisitically different or not. The formula for calculating your t value is in the lab 2 slides. You just need to plug in the values for the means, varience and sample size. You then compare your calculated t-value with the critical t-value in the t-table. You find your t-critical at the intersection of your chosen p-vlaue and the degrees of freedom (n-2). We set p at 0.05 (that is if we have significant result this is only likely to have at 5% likelihood occurred by chance). Then look down that column to the row that matches your n-2 (so if you had 30 leaves go to row 28). If t-calculated is > t-critical your result is significant (reject null hypothesis) If t-calculated is < t-critical your result is not significant (fail to reject null hypothesis) When you write about this in your report you would say something along the lines of: There was no difference in the mean amount of herbivory on kawakawa leaves at the tip and base of the branch (fig. 2). This was not significant at p=0.05 (t=XXXX). Nick has made a short video as background to the hypothesis. The idea is that young leaves are preferred by caterpillars so have higher levels of grazing.