Percentages and statistical significance


New Member
This might be a deceptively easy question, in which case I'm sorry.
In an experiment involving word lists and critical lures (i.e. recreation of Roediger and McDermott, 1995), participants were either given a distractor task or asked for immediate recall of word lists. We found that a high percentage of false memories for words created were in the distractor condition. How do I test that this is statistically significant?


New Member
Hi Sh7,

I am not a stats expert but, perhaps, my comments would be useful to you.
If I understood properly, you want to compare two proportions (you do not need to transform them to %): The proportion of false memories in the condition "distractor" and the proportion of false memories in the non-distractor condition. In most cases, this can be achieved by cross-tabulating the data and calculating significance by means of the chi-squared test. I attach the link to a website where you will find a brief description for chi-square and other alternative procedures: