How to Determine Sample Size With Known Power & Effect Size

In designing a research study, it is important to calculate how many participants, or what sample size, you will need in order to achieve statistical significance in your final results. Planning the level of power or effect you would like to achieve ahead of time, or a priori, will allow you to decide how large your sample size needs to be. Calculators that will allow you to do the work accurately and quickly can be located online.

Things You'll Need

  • Desired effect size
  • Desired statistical power level
  • Defined experiment type
Show More

Instructions

  1. Calculate Your Sample Size

    • 1

      Visit a website has a free statistical calculator. Click on the "Statistical Calculators" link on the page.

    • 2

      Select "Sample Size" from the list of available calculators in the green rectangular box.

    • 3

      Determine what type of analysis you will run on your data and select the appropriate link under "Sample Size." Follow the steps listed below depending on your analysis type.

    Hierarchical Multiple Regression

    • 4

      Click "Hierarchical Multiple Regression" if you are comparing the contribution of multiple sets of independent variables to predict change in your dependent variable.

    • 5

      Enter your desired alpha level, otherwise known as p-value or type I error rate. The default value is set to the generally accepted level of 0.05.

    • 6

      Enter the number of predictors from each set of your regression equation in the "Predictors in A" and "Predictors in B" boxes.

    • 7

      Enter your desired statistical power level.

    • 8

      Enter your expected effect size, using the guidelines provided. Click "Calculate" and scroll down to view the "Minimum Required Sample Size."

    Multiple Regression

    • 9

      Click "Multiple Regression" if you are using multiple independent variables to predict change in your dependent variable.

    • 10

      Enter your desired alpha level, otherwise known as p-value or type I error rate. The default value is set to the generally accepted level of 0.05.

    • 11

      Enter the number of predictors in your regression model, not including the regression coefficient.

    • 12

      Enter your desired statistical power level.

    • 13

      Enter your expected effect size, using the guidelines provided. Click "Calculate" and scroll down to view the "Minimum Required Sample Size."

    Student's t-Test

    • 14

      Click "Student's t-Test" if you are comparing means between two groups or between pre- and post-scores.

    • 15

      Enter your desired alpha level, otherwise known as p-value or type I error rate. The default value is set to the generally accepted level of 0.05.

    • 16

      Enter your expected effect size, using the guidelines provided.

    • 17

      Enter your desired statistical power level.

    • 18

      Click "Calculate" and scroll down to view the "Minimum Required Sample Size."

Learnify Hub © www.0685.com All Rights Reserved