How to Write a Results Statement for a T-Test or an ANOVA

Whether you're writing up your own research or simply taking a class on statistics or research methods, you may be called upon to write a results statement for your hypothesis test, using standard APA reporting conventions. This article explains how to do just that. Details are provided to assist those using the popular SPSS statistical software suite (now also known as PASW Statistics), but it is not necessary to have this program to benefit from the information in this article.

Instructions

1. For a t-test: You'll have to report the statement "t([degrees of freedom]) = [t-score], p is less than [significance value]." (for example, t(20) = 4.56, p is less than .01).

   For users of SPSS: In the SPSS output, SPSS reports the number of degrees of freedom in the column labeled "df", the t-value in the column labeled "t", and the significance value in the column labeled "Sig. (2-tailed)." (Do NOT use the value in the column that just says "Sig." ... this is the significance value for Levene's test, a totally different test that SPSS reports in the same table just for kicks.) There are a few other things to keep in mind if you get a nonsignificant result or are doing a one-tailed test, mentioned in the "Tips" section at the bottom of this article.Effects" table, df_between is the value up top, and df_within is the value in the error row.

2. For an f-test like an ANOVA: you'll want to report "F([df_between], [df_within]) = [f-score], p is less than [significance value]."

   For users of SPSS: For a oneway ANOVA, look at the "ANOVA" table; for a repeated-measures ANOVA, look at the "Test of Within-Subjects Effects" table. SPSS reports the f-value in the column labeled "F", and the significance value in the column labeled "Sig."

   For a one-way ANOVA, the table also clearly lists the degrees of freedom between and the degrees of freedom within.

   For a repeated-measures ANOVA, looking at the "df" column in the "Test of Within-Subjects Effects" table, df_between is the value up top, and df_within is the value in the error row.

3. Don't have a significant result to report? If your p-value is greater than .05, then there is no significant difference, and the last part of your statement should read "n.s., p is greater than .05". "n.s." stands for "not significant."

4. Insert your statement into the Results section of your paper or report. All done!

Tips & Warnings

• Remember that if you're doing a one-tailed (directional) test, you'll want to divide your 2-tailed significance value (p-value) in half.

• For users of SPSS: Even if SPSS reports that we have significance at some non-round number like the p = 0.00349 level, we usually report significance either the p less than .05, p less than .01, or p less than .001 level, whichever gives us the tightest bound. So if SPSS had 0.00349 in the significance column, we would usually report significance at the p less than .01 level.