I was hoping I could have some advice on the following;

I am trying to show a correlation between data;

I have 45 individual data values that range from 4.07 to 6.79. The range of this data is 2.69 but the distribution is random.

I am trying to show how the initial number correlates to a percentage change in correctly identifying a word from 1 time point (t2) to another (t4).

I have tried representing this by working out the mean of the lowest 15 numbers, the middle 15 numbers and the highest 15 numbers. I have then worked out the mean percentage change for the lowest 15 data vales and so on.

I have then plotted the data on a bar graph with the 3 categories on the x axis and percentage change on the y axis.

My worry is that as the data is distributed randomly this is very inaccurate and would I just be better off plotting all 45 data sets on a scatter graph.

Is there a better way of representing and analysis this data.

Many thanks for your help.