**Data Analysis Plan**
To model the effect of text difficulty on reading time: mixed effects linear regression.
To model the effect of text difficulty on mind wandering: mixed effects logistic regression with text difficulty as a fixed effect.
To address concern that increased mind wandering in the difficult condition might be confounded by longer reading times, we will do the following:
- compare mean reading times of the 4 easy texts and the 4 hard texts with a paired samples t-test.
- calculate any correlation between overall reading time and proportion of mind windering.
To model the effects of mind wandering on reading time: mixed effects linear regression on the presence and absence of mind wandering
To model the effects of mind wandering on comprehension: logistic regression
To model whether mind wandering moderates the effect of text difficulty on comprehension: regression comprehension scores on mind wandering separately for the easy and difficult texts.
**Results**
The participants in the difficult condition received lower scores than in the easy condition. This suggests that the manipulation of the text difficulty was successful χ2(1)=6.89, p < 0.01.
A mixed-effects linear regression model using text difficulty to predict sentence reading times was insignificant F(1, 2,061) = 0.294, p = 0.73. This contrasts the original study which found it to be significant. There was an increase of reading times (in milliseconds) in the difficult sentences (M = 7,801, SD =3,906) versus the easy sentences (M = 7,540, SD = 3,883).
To find out whether participants were more likely to mind wander with a difficult sentence as compared to an easy sentence we used a mixed-effects logistic regression model. χ2(1)=5.37, p<0.05. This result is in accordance to the original study.
To investigate whether mind wandering predicted reading times, we used a mixed-effects linear regression model. The results were as follows: F(1, 2,114) = 25.57 p < 0.01. Mind wandering predicted reading times after controlling for condition F(1, 2,112) = 26.65, p < 0.01.
We investigated whether mind wandering moderated the effect of text difficulty on reading times by including the Mind Wandering x Text Difficulty interaction term as a fixed effect. This approached a significant effect F(1, 2,069) = 8.69, p = 0.06.
A logistic regression model showed that participants were less likely to choose correctly on a comprehension question if they had mind wandered while reading the previous sentence χ2(1) = 28.31, p < 0.001.
