To Eat or Not to Eat
Gwen Morse
University of the Rockies

To Eat or Not to Eat
Introduction
One in five seniors have an income below 150 percent of the federal poverty line. (Food Research and Action Center, 2010). This creates a serious issue for seniors, often having to choose between eating and buying medications or paying bills. Many have no idea they may qualify for SNAP (Supplemental Nutrition Assistance Program), the official name of the food stamp program. If they are aware they may qualify, they are often too embarrassed to apply. According to the last census completed in 2010, Florida has the highest senior population in the United States at 17.3 percent. The State of Florida Department of Elder Affairs (2010) indicates the retirement industry is the second largest economic factor in the state. That being said, is there a difference between the number of retired males and retired females in Florida receiving SNAP benefits?
Hypothesis
Hypothesis1= There is a difference in the number of retired men and retired females receiving SNAP benefits.
Hypothesis0=There is no difference between the number of retired males or retired females in Florida receiving SNAP benefits.
Variables
The dependent variable is receipt of SNAP benefits in Florida. It is categorical, criterion, discrete, and nominal. The independent variable questions if more retired females or retired males in Florida are receiving SNAP benefits. It is categorical, predictive, discrete and nominal.
Methodology
There are two ways to collect data in this case. The first would be to use data supplied by the Florida Department of Elder Affairs and SNAP. The second would be by random survey of retired persons, supplied by the local area AARP Association, or some other local agency dealing with senior issues, such as the Senior Resource Association or a local senior center. Either results would provide information regarding whether more men or woman receive benefits. If using data supplied by the state agencies, the data sample would be very close to the actual population. This would reduce the possibility of error and bias that could occur when conducting a survey. If using a survey, the sample would still need to be should be relatively large, like 2000 or 3000, in order to attempt to reduce error and bias. Error can easily happen when using the survey if the data came from married couples or other relationships where seniors are living together or with family, and receive SNAP for the entire family, or there are duplications.
Chi Test and Goodness of Fit This study lends itself to the chi square test and Goodness of Fit because it examines two unrelated variables. In order to test a null hypothesis, it is necessary to first know the total number of seniors receiving SNAP benefits, and then divide it by two (male and female). The expected number would be equal. The observed number would be actual number of males and the number of females as supplied by the agency data or survey data. The observed number is subtracted from the expected data (the equal number in each category). The difference is squared and divided by the expected. If the observed gets close to the expected, the chi squared approaches zero, and the null hypothesis is accepted. If that number is larger, we will reject the null hypothesis. Since df (Degrees of Freedom) = Categories (2)