Actually SPSS has not the power to calculate directly for determining sample size. But there are some separate (expensive) products can be used for this purpose. There are a number of very useful websites exit for power/sample size calculations that you can find with Google. Many of these have free calculators. Also, power/sample size are design specific. You need to have an idea of how you will analyze the data, size of the effect you want to find, the variation in data, the level of significance and the population size as well. Sample sizes are needed in statistics because it is not always possible to conduct an analysis of a whole population. In that case, SPSS can help you to choose your sample in a snap. I am describing the process below.

1. Firstly, select the ̔Data̕ menu and then click ̔Select Cases̕.

2. Check the ̔Random sample of cases̕ button, then check the ̔Filtered̕ button.

3. Click ̔Sample̕ in the center of the dialog box, then check the ̔Approximately̕ button.

4. Type a percentage into the box. For example, type ̔5%̕ if you want to sample 5 percent of the population.

5. Click ̔Continue̕ and then click ̔OK̕.

This completes the selection of a random sample. Similarly if you know exact how many cases you want to select (the sample size) then click the ̔exactly’ button and write the exact number that you want and then ̔cases from the first̕ from how many cases you want to choose the sample that you need to mention. Then click ‘Continue̕ and then click ‘OK̕.

To check that the system has picked a representative sample:

1. Run a descriptive output for the variable.

2. Click ̔Analyze̕ and then click ‘Frequencies̕.

3. Click ̔variable’ in the left window of the dialog box, then click the arrow key in the middle to transfer ̔variable to the right-hand window.

4. Click ̔Statistics̕, then place check marks in ̔Mean, Std. Deviation̕ and ̔S.E. mean̕.

5. Click ̔Continue̕ and then click ̔OK̕.

This will deliver a report that will tell you if you have chosen a representative sample. If you haven’t, run through the procedure again with a different percentage in your SPSS research.

When doing an SPSS research analysis, the following factors should be considered when deciding on how big your sample size should be:
1. Purpose of the study – if your SPSS research entails doing a segmentation of the market or building predictive models via SPSS’ multivariate techniques (e.g. factor analysis, regression analysis, cluster analysis), you will need a bigger sample size. The acceptable ratio of the number of cases or respondents to be included in the study is 20 times the number of variables to be included in the multivariate analysis.
2. The available population size or sampling frame also plays an important role in deciding on how big or small your sample size should be.
3. One should also consider the level of precision to be used. This is also called the margin of error. The normal margin of error for any study is at 10%. Lower the margin would require larger sample size.
4. Aside from the margin of error, one should also consider the degree of variability to be assumed in the study. A normal study uses a variance of 0.5.
5. Time and cost constraints are also important factors to be considered when deciding the sample size for any SPSS research.

Actually SPSS has not the power to calculate directly for determining sample size. But there are some separate (expensive) products can be used for this purpose. There are a number of very useful websites exit for power/sample size calculations that you can find with Google. Many of these have free calculators. Also, power/sample size are design specific. You need to have an idea of how you will analyze the data, size of the effect you want to find, the variation in data, the level of significance and the population size as well. Sample sizes are needed in statistics because it is not always possible to conduct an analysis of a whole population. In that case, SPSS can help you to choose your sample in a snap. I am describing the process below.

1. Firstly, select the ̔Data̕ menu and then click ̔Select Cases̕.

2. Check the ̔Random sample of cases̕ button, then check the ̔Filtered̕ button.

3. Click ̔Sample̕ in the center of the dialog box, then check the ̔Approximately̕ button.

4. Type a percentage into the box. For example, type ̔5%̕ if you want to sample 5 percent of the population.

5. Click ̔Continue̕ and then click ̔OK̕.

This completes the selection of a random sample. Similarly if you know exact how many cases you want to select (the sample size) then click the ̔exactly’ button and write the exact number that you want and then ̔cases from the first̕ from how many cases you want to choose the sample that you need to mention. Then click ‘Continue̕ and then click ‘OK̕.

To check that the system has picked a representative sample:

1. Run a descriptive output for the variable.

2. Click ̔Analyze̕ and then click ‘Frequencies̕.

3. Click ̔variable’ in the left window of the dialog box, then click the arrow key in the middle to transfer ̔variable to the right-hand window.

4. Click ̔Statistics̕, then place check marks in ̔Mean, Std. Deviation̕ and ̔S.E. mean̕.

5. Click ̔Continue̕ and then click ̔OK̕.

This will deliver a report that will tell you if you have chosen a representative sample. If you haven’t, run through the procedure again with a different percentage in your SPSS research.

When doing an SPSS research analysis, the following factors should be considered when deciding on how big your sample size should be:

1. Purpose of the study – if your SPSS research entails doing a segmentation of the market or building predictive models via SPSS’ multivariate techniques (e.g. factor analysis, regression analysis, cluster analysis), you will need a bigger sample size. The acceptable ratio of the number of cases or respondents to be included in the study is 20 times the number of variables to be included in the multivariate analysis.

2. The available population size or sampling frame also plays an important role in deciding on how big or small your sample size should be.

3. One should also consider the level of precision to be used. This is also called the margin of error. The normal margin of error for any study is at 10%. Lower the margin would require larger sample size.

4. Aside from the margin of error, one should also consider the degree of variability to be assumed in the study. A normal study uses a variance of 0.5.

5. Time and cost constraints are also important factors to be considered when deciding the sample size for any SPSS research.