This is probably the most overwhelming 'writing' component of the paper for students. However, there are some simple and basic rules for writing results sections that, if followed, make it potentially the easiest writing because it is so technical. Some main points are:

• Organize results in a logical sequence
• Use headings, if appropriate
• For every statistic, be sure that you report the statistic (italicized), the degrees of freedom, the value, and the p (italicized) level, set off from the sentence by a comma (not parenthesis). See the APA manual for examples.
• For significant t-tests and Anovas, you must report means and interpret the means. For example: "An ANOVA was conducted with Attachment Style (secure, anxious, avoidant) as the independent variable and Self-esteem as the dependent variable. A significant main effect was found, F(2,43) = 4.56, p < .05. The means indicated that respondents with a secure attachment style scored significantly higher on self-esteem (mean = 3.45) than those with an anxious (mean = 2.98) and avoidance (mean = 2.56) attachment styles."
• For correlations, you must interpret the direction of the relationship. For example: "A significant correlation was found between self-esteem and internal-external locus of control, r (42) = .32, p < .05, indicating that internal respondents reported higher scores on self-esteem."
• Most means, standard deviations, statistics, and p-values will be rounded off to two decimal points
• Only use a table if there are many statistics (e.g. bivariate correlations, multiple dependent variables for tests of significant differences) that are more efficiently presented that way. Otherwise, report statistics in text. Tables are not embedded in the text. See APA manual for format and locations for Tables.
• Only use figures if you need to report a significant interaction effect for an ANOVA. See APA manual for format and location of Figures.

Below are some headlines you can use to structure your results section and the content that can be included in each section:

Descriptive information:

Discuss if something in the demographic is interesting, such as the fact that a high percentage of students worked, or were married (that would be interesting because college students traditionally don't work or have other responsibilities), if not, just mention and put in Table 1 (and mention table 1). You need to think about your sample, etc. If you were the reader, what of all the data that you have would make a good story, would change their perceptions, would help them understand your findings?

In this section is also where you give descriptives of the variables. Such as In the present sample, subjects scored relatively high on depressive symptoms as measured by the CESD (M = 17.67, SD = 3.45, Range: 3 to 43). Notice that by mentioning the scale, you save the reader the effort of going back to measures section to remind himself of what the measure for depression was. On the other hand, ratings for anxiety were within the normal range for the STAI (put in numbers as above). This information situates the reader to interpret your findings. Depending on your question, a group that scores higher than average on a variable will answer the question differently than one that scores average, for example.

Correlation:

Usually correlations are not used to answer questions. In fact, 'correlational studies' are considered less than desirable because they only point to relationships and not much else. For some of you, the correlation is the answer to the main question and that is appropriate at this level. Regardless, usually, the correlation follows the demographic information.

Correlations between all the variables in the study are presented in Table 2. As expected, the subscales of the XXX correlate moderately. Of note is the correlation between anxiety and credit hours (r = .45, p = .034) and the negative correlation between anxiety and class standing (r = .34, P = .012). These suggest that higher anxiety is associated with taking more units per semester and to being a lowerclassman rather than an upperclassman.

If there are no noteworthy correlations, you state so.

Main analysis:

You should start with the main analysis for the main question (unless correlation was the main analysis). It won't hurt to remind your reader what your hypothesis was. You should always think what is most convenient to the reader. Therefore, you can say: the hypothesis that men would score higher than women in anxiety was tested with an analysis of variance. (notice how you can say that this is the hypothesis and you say what analysis you did to test it). You need to be specific in the results section. Don't suppose the reader knows what you did. Everything you did should be spelled out, how, with whom, in what order. If the tests are not significant, then you follow the previous sentence with There were no significant differences between groups (p > .05). If there were differences, you follow up with Men scored higher than women, F(df) = number, p = ; (see Table 2) Notice that you only give statistics for significant findings. People who understand stats will be able to evaluate the value of the numbers you give. Any significant differences should have the means presented, either within the text if there are only a couple or in a separate table if there are several.

Related to that, even if your means look very different, if an ANOVA didn't find differences, you cannot say the groups differed. Remember probability? If ANOVA is not significant, there is a high probability that the differences you see could have happened by chance and never happen again!

Do the same with any other analysis you did. If it was not the main question, then say what you did, such as An anova was run to examine differences in sleep between various ethnic groups.

Don't generalize (we conducted several analysis) or group things together. The results section is a technical report. It should be detailed, specific. If you run an analysis, the IV, group, and DV should be clearly spelled out (see the examples above).

Regressions are reported as follows:

A regression was run to examine the relationship between anxiety as the dependent variable and mood and self-esteem as the predictors. The regression equation was not significant (P > .05). That would be it for non-significant findings. For significant findings:

The regression equation predicted 6% of the variance in anxiety. Both mood (B = .56, p < .001) and self-esteem (B = .23, p = .013) entered the regression equation. If only one entered, state the one and make it clear that the other didn't.

As with all the sections you write, you will do well to ask a classmate to review a well-developed draft. This person can give you feedback as to whether you are clearly presenting the findings or leaving things out supposing your reader knows as much as you do (one of the most common mistakes capstone students make in writing).