deletions | additions       diff --git a/e1_results.tex b/e1_results.tex  index c9a1da0..d35a462 100644  --- a/e1_results.tex  +++ b/e1_results.tex 
...
\subsection{Results}  \subsubsection{Data pre-processing}  All analyses were performed using R \citep{R2009}. \citep{R2015}.  All trials during which a response was made when there was no target on screen (1.7\% of trials overall) were removed. Participants missed targets on less than 1\% of trials overall, therefore an explicit analysis of miss rates was not possible. Trials with response times (RTs) faster than 200ms ($<$1\%) or slower than 1000ms ($<$1\%) were removed from the analysis. \subsubsection{Response time}  Trial-by-trial RTs were submitted to the mixed effects modelling approach detailed above, yielding a $LR$ for the effects of SOA, Cue Validity and their interaction of 1.071 B, 58.23 B, and 36.09 B, suggesting strong evidence for all effects. Analysis of the data excluding the neutral condition obtained similar evidence for the effects of Cue Validity (46.81 B) and the SOA by Cue Validity interaction (27.16 B), but diminished evidence for a main effect of SOA (-0.52 B). Figure \ref{exo_rt} depicts this data, suggesting that while there were no differences between cue types at the early SOA, the typical pattern of IOR (valid slower than invalid) manifests at the later SOA. RT results here  \subsubsection{Probability of memory}  Probability of memory results here.  \subsubsection{Fidelity of memory}  Fidelity of memory results here.