Krisanna Machtmes, PhD

Evaluating an extension program allows us to share our results with stakeholders and program participants which is a primary method funding justification of cooperative extension programs. The evaluation results indicate the impact our programs have on the lives of our participants. We want to write up the results of these evaluations in such a manner the reader is informed of the results, who has participated in our program, and how their lives were changed for the better due to their participation.

Many individuals struggle with understanding how to analyze and report program evaluation data correctly. We collect a variety of data on surveys ranging from demographic data (age, gender, etc) to scale data (strongly disagree, strongly agree) and open ended questions as well.

Data analysis begins with knowing how many individuals attended our program. This is referred to as our sample size. The sample size number is used to calculate many of the numbers we will be reporting.

Analyzing demographic data such as gender, race, ethnicity, age, income, socioeconomic status, and a description of where people live (such as rural or urban) requires us to know whether the data is measured as the following: nominal, ordinal, interval, or ratio data.

Most of the demographic data that is collected is nominal, ordinal, or ratio data.

**Nominal data**is data that has no natural order to it and is used to classify a response variable such as gender. For example, a program participant will be either a male or female. This type of data can only be reported as frequency data, such as stating 30% of the participants were male. To calculate frequency data, you count the number of times the data occurs in a category and then divide by the total number of participants in the program. Data that is nominal cannot be reported as means and standard deviations as you cannot have 3.4 males.

**Ordinal data**is data that has a particular order to the data such as elementary school, junior high school, and high school education levels. Ordinal data can be reported as frequency data so that we can say the following: 20% of the schools were elementary schools. Data that is ordinal cannot be reported as means and standard deviations as you cannot have 4.6 elementary schools.

**Interval data**is the typical data that we collect in a survey and/or program evaluation. For example, we often ask participants if the program was poor or fair or good. This type of data may be reported as means and standard deviations.

**Ratio data**is data with a true zero, such as the height of a tree or the age of a person. This data may also be reported as means and standard deviations. You can ask a person their age at their last birthday, and this type of question would result in ratio data. You would report this data as means and standard deviations.

*Calculating a Mean and Standard Deviation *

An example: Forty people have completed your financial educational program survey. You asked all of the people to list their current age on this survey.

To calculate the mean age of the respondents (respondents are people who complete your survey and give you usable data) you will add all of the ages up and then divide the total by the number 40 (because 40 people completed your survey). This can easily be done on a calculator or using an Excel spread sheet (see the additional information below on how to use excel to analyze program data).

A standard deviation for the age will inform you about the spread of your numbers about the mean value. The spread refers to how close and far all of the age values are to your calculate mean, thus, you are able to see if there are extreme values.

When examining the mean number you must you caution because the mean or average score is influenced by extreme low or high numbers – so if you have someone in your age group that is 89 and the next highest age is 40 – that age of 89 will influence your mean age.

Additional information for calculating and analyzing demographic data

See the following websites for more information – Excel is commonly used to analyze program data – here is some additional information regarding using Excel: http://learningstore.uwex.edu/Assets/pdfs/G3658-14.pdf

Additional information for analyzing survey data is available at this pdf

http://learningstore.uwex.edu/Assets/pdfs/G3658-06.pdf

This is very helpful information.