Skip to main content

Greg McVerry

Writing Up Regression

Social support significantly predicted depression scores, b = –.34, t(225) = 6.53, p < .001. Social support also explained a significant proportion of variance in

2 min read

Social support significantly predicted depression scores, b = –.34, t(225) = 6.53, p < .001. Social support also explained a significant proportion of variance in depression scores, R2 = .12, F(1, 225) = 42.64, p < .001.

This is from the APA. I don't usually care about if it predicted. I want to know the amount of variance explained and the effect size.

In the first sentence they tell you the model worked and it's significance. The second sentence includes the variance explained (my fav dish). They also include a second clause with the R-squared which is a measure of effect size.

You should really just include a table because all these numbers break reading comprehension. In the end do what the editor says.

Resultsts of the linear regression, as presented in Table 4.11, indicated, first that verbal intelligence, as measured by a vocabulary test, explained 16.9% of the variance, which was significant, F(1,93) = 20.530, p<.001. After verbal ability was accounted for, the three predictor variables of background knowledge, evaluation skills, and scores on a self‐report measure of dispositions of online reading comprehension explained an additional 22.4% of the variance on the scores of the less restricted ORCA, which was significant F(1,92) = 15.415 p<.001. In the full model, scores on a critical evaluation measure explained a significant amount of unique variance (β= .335 p< .001). Background knowledge also contributed significantly to the prediction of scores on the less restricted ORCA (β= .211 p= .021). Dispositions of online reading comprehension did not contribute significantly to predictions in the model (β= .117 p= .162). The addition of the variables in the second step of the less restricted model had a large effect, Cohen’s F= .358.

This is my dissertation version. It is way too long. Do not write regression results like a dissertation. Use a table.

CLMOOC

Prev | Home | Join | ? | Next