ii Exercise C: Explain what's wrong here and improve it [more of an essay question you may provide hypothetical R-code to support your argument, but you do not have to run it on actual data) The following is an entirely hypothetical example. [Unfortunately, as a reviewer ofjournal articles which report mixed effect model analyses one encounters cases like this alarmingly frequently...] Imagine a researcher ran an experiment to test whether Predictor A and Predictor B interact in predicting occurrences of 'correct' (1) versus 'incorrect' (0) responses in mathematical problem solving. Both predictors are categorical and have two levels each. Moreover, in the experiment, both predictors are withinsubjects and within-items (repeated-measures design with multiple trials per design cell). For analysis, the researcher codes the two levels _ maybe deviation coding is better? _ _ _ of each predictor as 0 and 1, respectively. They run the followmg \"mixed effects model with random effects for participants and items, using the R-package lme4\" (quoted from their imaginary report): lmer' correct A B participant data math_data The researcher reports that \"in the model summary, the expected interaction did not come out significant (p = .07), but interestingly, the main effect of B was significant at p