48.Aspx Briefly skim this article on how to report data with small numbers (pay attention to pp 2-6). Review the guidance issued by the Federal Trade Commission on protecting personal data. Watch this Podcast: Privacy and Predictions
What did you find most interesting related to the use of data in the articles and podcasts?
Royse, Thyer, and Padgett discuss privacy on pp. 44-47 and outcome evaluations on pp 228-234 (chapter 9). Pick one element of their privacy discussion and briefly describe it focusing on how it applies to an evaluation of program outcomes. Without getting into specifics, how well does your agency protect the identities of people included in its data?
How useful are descriptive statistics to a program evaluator when the sample size is very small?
Full Answer Section
When the sample size is very small, descriptive statistics can be less reliable and informative. This is because they are more likely to be influenced by chance outliers. Additionally, small sample sizes can make it difficult to generalize the findings of the evaluation to the larger population.
Despite these limitations, descriptive statistics can still be useful to a program evaluator when the sample size is very small. For example, descriptive statistics can be used to:
Identify trends and patterns in the data. Even with a small sample size, it may be possible to identify trends and patterns in the data that suggest that the program is having a desired effect.
Develop hypotheses for future research. Descriptive statistics from a small sample study can be used to develop hypotheses for future research studies with larger sample sizes.
Compare different groups of participants. Descriptive statistics can be used to compare different groups of participants in a small sample study, such as participants who did and did not receive the program intervention.
Provide context for qualitative data. Descriptive statistics can be used to provide context for qualitative data, such as interview data or focus group data.
Here are some tips for using descriptive statistics with small sample sizes:
Use multiple measures. Using multiple measures of the same concept can help to reduce the impact of chance outliers.
Use nonparametric statistics. Nonparametric statistics are less sensitive to the assumptions of normality than parametric statistics.
Report confidence intervals. Confidence intervals provide a range of values that is likely to contain the true population parameter. This information can be helpful to readers in interpreting the results of the evaluation.
Be cautious about generalizing the findings. Remember that the findings of a small sample study may not be generalizable to the larger population.
Here are some examples of how descriptive statistics can be used in program evaluation with small sample sizes:
A program evaluator might use descriptive statistics to compare the mean pre-test and post-test scores of a small group of participants in a literacy program. If the mean post-test score is significantly higher than the mean pre-test score, this suggests that the program is having a positive effect on participants' literacy skills.
A program evaluator might use descriptive statistics to compare the percentage of participants in a job training program who are employed at the end of the program to the percentage of participants who are unemployed at the end of the program. If the percentage of employed participants is significantly higher than the percentage of unemployed participants, this suggests that the program is helping participants to find employment.
A program evaluator might use descriptive statistics to compare the different types of feedback that participants in a customer service training program provide about the program. This information could be used to improve the program in the future.
Descriptive statistics can be useful to a program evaluator when the sample size is very small, but their limitations should be kept in mind.
Descriptive statistics are used to describe the characteristics of a sample. They can be used to calculate measures such as the mean, median, mode, standard deviation, and range. These measures can provide information about the central tendency, variability, and distribution of the data.