This question is about how statistics are reported in research articles. Below i
ID: 3224457 • Letter: T
Question
This question is about how statistics are reported in research articles. Below is a brief excerpt from a study looking at how trained visual artists differ from non-artists in copying photographs of faces by sticking small pieces of tape within a grid to create pixelated-looking drawings of faces. The researchers were interested in whether the two groups differed in the accuracy of their depictions of faces. The dependent variable was a measure of accuracy, in which higher values indicate greater accuracy in copying the face. A total of 18 artists and 27 non-artists participated. Read the excerpt and answer the following questions. The excerpt: A comparison of the accuracy values of artists and non-artists revealed a highly reliable difference, t(43) = -6.97, pExplanation / Answer
A. Null hypothesis H0: µ1=µ2 is to be tested against the two tailed alternative hypothesis H1: µ1µ2
B. Let Artist denote the variable which takes two values A and N for artist and non-artist respectively, then Artist is the independent nominal dichotomous variable
C. Two-sample t-test was used and as degree of freedom=n1+n2-2=43, so equal variance is assumed for the test.
D. The sampling distribution used was t-distribution with degree of freedom=n1+n2-2=43.
E. Null hypothesis was rejected as p-value is less than 0.05.
F. Effect size is given to be 2.15 which is a very large effect size.
G. We observe that the artist mean accuracy 1.76 is higher than non-artisits mean accuracy of 0.96, so artist are more efficient in performance as compared to non-artist. This results is statistically significant as p<0.001.
H. Power should be high because we can detect small difference as effect size is very large. Power and effect size are consistent as sample size is large.