Office of Academic Assessment

How many students do I need to assess?

The number of students you need to assess, depends the answers to four very simple questions. These four questions are: What is the size of your population (N)?  How confident do you need to be about your finding? How accurate does your finding need to be? What is your target performance? I know this looks very mathematical and statistica.  Don’t worry there is an easy way to simply fill in each blank and all of the calculations will be performed for you

N = (population size)

z  = (confidence level)

E  = (+- error)

p  = target performance

Population Size (N) This is the size of the population you are trying to measure. This number will usually be the total number of students in your program at the level you are attempting to assess.

Confidence Level (z) is the value that specifies the level of confidence you want in your data fall within the range described. Typical levels of confidence are 95%, in which case the z is set to 1.96. the z is set to 2.56 for 99% confidence interval and z = 1.67 for a 90% confidence interval.

Error (E) Is the percent of error we are willing to accept in our statistic. We are all used to seeing national surveys in which the percentage is given as plus or minus a certain percentage. This indicates how much inaccuracy in the true score we are willing to tolerate.

Target Performance (p) This is the level of performance that you are expecting your sample to demonstrate.  For example if you have set the criteria for 80% of students to achieve minimum performance then p = .8 If you set the criteria for 95 percent of students to achieve the criteria then p = .95

Fill the green portion of the attached spreadsheet to calculate the appropriate sample size

Sample Size Calculator

 

Source Document

http://uregina.ca/~morrisev/Sociology/Sampling%20from%20small%20populations.htm