showed a trendp = 0.09
A significant interaction between valence and evaluation time ( F (1,252) = 4.46, p = 0.036) showed that ratings for later positively and negatively paired Fribbles did not differ at baseline ( p = 0.229), but showed a trend towards a more positive evaluation of positively paired Fribbles compared with negatively paired Fribbles ( p = 0.09, p corr = 0.364) at the delayed rating.