CHAPTER 3 Methodology Survey Design

For the purposes of this project, the University of California Cooperative Extension’s Emergency Food Client Questionnaire (UCCE-EFCQ) and the United States Department of Agriculture’s Core Food Security Module (USDA-CFSM) were combined so that a more in-depth analysis of factors influencing hunger and food insecurity could be conducted. One survey was developed from these two models and revised to meet the specific needs of Calaveras County. This revised version was called the “Client Survey” (Appendix B, pp. 84-99). Questions 1 through 32 on the Client Survey were taken from the UCCE-EFCQ. Questions 33 through 48 on the client survey were taken from the USDA-CFSM. The following list outlines the changes to the original UCCE-EFCQ that were made and the reasoning behind those changes:
Revised “Client Survey”(p. 85)–Question 6: What kind of housing do you have?

Responses were worded to more specifically differentiate between house, mobile home, and recreational vehicle (RV). For example, “own home” on the original survey was changed to two separate responses “own house (not mobile home)” and “own mobile home” on the revised “Client Survey”. The response “mobile home/RV” on the original survey was changed to two separate responses “rent mobile home” and “RV”. These changes were based on the knowledge that Calaveras County low-income residents might choose to live in mobile homes or recreational

vehicles. If a large percentage of clients was found to choose this lifestyle,

intervention programs could be designed to target this particular subgroup.

Revised “Client Survey”(p. 86)–Question 9: What are the ages of the people who live in your house, apartment, etc. not including yourself? The question on the UCCE- EFCQ that asked for information on the ages of children and adults in the respondent’s family seemed confusing and complicated. The question on the UCCE- EFCQ was redesigned in an attempt to make it less complicated and more understandable to respondents.

Revised “Client Survey” (p. 88)–Question 16: What are you or your household’s monthly sources of income? (your best estimate) Two responses on the UCCE- EFCQ needed clarification. One response, Aid to Families with Dependent Children (AFDC), had recently changed its title to Temporary Aid to Needy Families (TANF); therefore, both acronyms were used to facilitate respondent understanding. The response “salary” was clarified by changing it to “Employment (Salary)”.

Revised “Client Survey” (p. 88)–Question 17: How much does your household spend each month on the following items? (your best estimate) The list of responses was made more specific. For example—“rent/mortgage” was separated on the revised survey to “rent” and “house or mobile home payment”. The response “gas/electricity” was changed to six separate responses “gas”, “electricity”, “wood”, “propane”, “sewer”, and “garbage” to reflect the variety of utilities used in Calaveras County.

19

20

Revised “Client Survey” (p. 90)–Question 23: Why do you or other members of your household shop there? This question refers to the type of store where respondents buy their groceries. Due to the rural nature of Calaveras County, the answer, “no transportation to get anywhere else” was added to the response choices.

Revised “Client Survey” (p. 90)–Question 25: Would any of the following things be helpful to you or other members of your household? One of the purposes of doing the survey was to identify the needs of Calaveras County residents who seek emergency food assistance. The addition of this question allowed the development of a set of responses to identify specific workshops, informational materials, or interventions that might help respondents manage their limited resources more efficiently.

Revised “Client Survey” (pp. 91-93)—Questions 26 through 31: The original UCCE- EFCQ had several questions related to soup kitchens and emergency food. All questions relating to soup kitchens were eliminated since soup kitchens do not exist in Calaveras County. The questions relating to emergency food recipients were grouped together on pages 91 through 93. The goal of these questions was to elicit details about respondent households that were utilizing the emergency food system in Calaveras County. The questions asked for more in-depth answers in areas such as where most households sought emergency food, the frequency of need, the frequency of needing but not receiving, the reason for need, problems getting to emergency food sites, and the quality of the food received. If a respondent had not received emergency food in Calaveras County, the first of these questions allowed a “no” response and the respondent was instructed to skip to question #33.

21

Revised “Client Survey” (p. 92)–Question 29: What problems did you have now or in the past getting to the emergency food site? One of the goals of the survey was to identify the obstacles and barriers respondents experienced when seeking emergency food assistance. This question was added to elicit more in-depth information.

Revised “Client Survey” (pp. 94-98)—Questions 33 through 48): The USDA-CFSM questionnaire was added to determine the food security status of the respondent household instead of food security questions 28 through 31 on the original UCCE- EFCQ.

Pilot Testing the Client Survey

Three family advocates from the Calaveras Head Start State Preschool volunteered to pilot test the Client Survey with parents in the preschool program. The purpose of pilot testing was: a) to gauge the amount of time needed to complete the survey; b) to determine if there would be any problems with the changes made to the survey as outlined above; and c) to make a qualitative judgment on the best method of administering the survey.

A training meeting for the family advocates was scheduled to discuss the purpose of the survey, the target population, the amount of time it would take to complete the survey, and the method of survey administration (see Appendix E, pages 105 to109, for Training Materials). The family advocates were asked to take the survey themselves and make recommendations based on their experience with parents. Family advocates suggested parents would want to read and answer the questions themselves with minimal intervention.

22

Based on this recommendation, several more changes were made on the Client Survey used in this study. The instructions to interviewers on the original USDA-CFSM were reworded or eliminated on the Client Survey to allow for easier reading by the respondent. The questions themselves remained exactly the same except that abbreviations such as “DK” was changed to “don’t know” and “R”, which means “refused to answer”, was changed to “prefer not to answer”. The response “I/We have no children in our home” was added to questions pertaining to households with children so households without children would have a response. The essence of the questions was not altered with these changes but the readability of the survey was improved.

A total of 60 surveys were given to three family advocates to administer at their preschool sites in Angels Camp, West Point, Arnold, Copperopolis, Valley Springs, and San Andreas, California. Family advocates were instructed to emphasize that they would be available if the parents had any problems understanding the survey. Twenty-six parents volunteered to complete the survey. Results of the pilot test were as follows: The survey took approximately 10 to 15 minutes to complete. The family advocates

determined this to be an acceptable amount of time.
Family advocates reported that, as expected, parents wanted to fill out the survey

themselves with minimal intervention. Some parents wanted to take it home and

return it the next day.
Parents verbally reported to the family advocates that they understood the questions

and were able to answer them with no problems. All questions that had been changed or added were answered and seemed to have no problems.

Sample Selection

23

A convenience sample was taken of Calaveras County residents accessing agencies, churches, and organizations serving households living at or below 185% of the poverty level (see Appendix A, Table 2, page 54 for 1999 Poverty Guidelines). This sample was appropriate for this project because food insecurity has been found to increase as households fall below 185% of the poverty level (Bickel, Hamilton, Cook, et al., 1997a, p. 53).

Data Collection Sites
Two primary criteria were used to identify agencies, churches, and organizations for

participation in survey administration. First, surveys were to be given at sites in Calaveras County that provided services for residents living at or below 185% of the poverty level. Secondly, sites were selected that were representative of the five main geographic areas of Calaveras County: (a) Angels Camp, Murphys, Vallecito, Douglas Flat, Altaville, Copperopolis; (b) Arnold, Avery, Hathaway Pines; (c) San Andreas, Mountain Ranch, Sheep Ranch, Mokelumne Hill; (d) West Point, Wilseyville, Railroad Flat, Glencoe; and (e) Valley Springs, Burson, Jenny Lind, Wallace. (see map in Appendix A, p. 50)

Once the agencies, churches, and organizations were identified, two methods were used to solicit participation in administering the survey. Individuals attending the Emergency Assistance Relief Services (EARS), the Children Alliance, and the Providers Network meetings were asked to volunteer to administer surveys to their clients. The second method was to ask for participation from social service agencies representing the

24

five main geographic areas of Calaveras County. For example, it was determined that parents utilizing Motherlode Women, Infant, & Children (WIC) access services at sites in Angels Camp, Arnold, San Andreas, West Point, and Valley Springs; therefore, surveys were administered to a sampling of clients at each WIC site.

Representatives from eleven emergency food programs, social service agencies, and government programs volunteered to participate in administering the survey. Three of these programs allowed researchers to personally administer the survey: Community Emergency Service Programs, Calaveras Works and Human Services, and WIC. Other participating agencies, organizations, and churches were: Child Care Resources; Calaveras Head Start State Preschool; SHARE; San Andreas Assembly of God; Sierra Aids Council; St. Vincent de Paul; Seventh Day Adventist Community Services; and International Chaplain’s Association.

Training of Community Volunteers

Training sessions were held to familiarize volunteers with the survey instrument and to facilitate continuity in answering any questions that might arise from survey respondents. A packet (Appendix E, pp. 105-109) was given to each volunteer which included: a) the purpose of the survey; b) the importance of the survey; c) definitions of food security, food insecurity, and hunger; d) a resource number to contact for emergency food services; and e) the sources of the survey questions.

Community volunteers giving the survey were instructed to ask their clients if they would voluntarily fill out a survey. If the site was an emergency food program or food voucher program, the volunteers were instructed to give clients their food bag or

25

vouchers prior to asking them to fill out the survey. Respondents were to be assured their answers would not be used to determine eligibility in any programs. Emphasis was to be placed on participation being confidential. The survey was to be self-administered but volunteers were asked to remain readily available to clarify any questions the respondent might have or to read questions if necessary.

Sampling

Volunteers asked 174 of their clients to complete the survey. Of the 174 surveys returned, 159 were completed correctly and used in the final analysis. Volunteers reported most clients self-administered the survey although the volunteers remained readily available for any questions or to read the questions if necessary. Volunteers were not asked to keep a tally of those who refused to participate but reported very few clients would not fill out the survey. Three factors seemed to be consistent for those few clients who did refuse to fill out the survey: a) refusal due to time constraints; b) discomfort with giving others personal data; and c) reluctance of their elderly clients to fill out the form or be interviewed.

Timeline
The survey was administered from May through October of 1999. The winter months

were intentionally avoided so that the possibility of higher rates of winter unemployment and/or increased expenses during the holiday season would not affect the data. The majority (54%) of surveys were administered in October 1999 with the remainder fairly evenly distributed between the other months (14% in May 1999, 6% in June 1999, 11%

in July 1999, 4% in August 1999, and 11% in September 1999). The large number of surveys administered in October was due primarily to the 72 surveys that were administered to WIC clientele.

Analysis Methods

26

The researcher’s question, “To what degree have households that utilize social services in Calaveras County experienced hunger in the past 12 months?” was answered using the responses to questions 34 through 48 on the Client Survey (Appendix B, pp. 95- 98). SPSS 9.0.0 for Windows was used to tally the number of affirmative responses to the questions as directed in the Guide to Implementing the Core Food Security Module5 (Bickel, Price, Hamilton, et al., 1997). The greater the number of affirmative responses given to questions 34 through 48 on the Client Survey, the more severe the household’s food insecurity and the higher the household’s food security scale value. The food security scale values ranged from 0 (being the most food secure) to 10 (indicating food insecurity with severe hunger). Based on the household’s food security scale value, each household was then be assigned a food security status level (0 = food secure; 1 = food insecure without hunger; 2 = food insecure with moderate hunger; and 3 = food insecure with severe hunger) according to where the household was on the scale. For purposes of this project, the scores of 2 (food insecure with moderate hunger) and 3 (food insecure with severe hunger) were considered one category called “food insecure with hunger”. _________________

5. The Guide to Implementing the Core Food Security Module can be downloaded from the following site: http://www.ers.usda.gov/briefing/foodsecurity/surveytools/index.htm/Guide to Measuring Household Food Security.

27

Appendix D on page 103 of this thesis project shows how the total number of affirmative responses on questions 34 through 48 of the Client Survey corresponds to a food security scale value, which can then be converted to a food security status level.

SPSS 9.0.0 for Windows (1998) was used to conduct interpretive tabular analysis (frequencies, percents, means) and cross-tabular analysis (number of respondent households with a specific food security status level cross-tabulated with a specific sociodemographic characteristic) to determine the answer to the following questions: What are the relationships between the degree of food insecurity and specific

sociodemographic characteristics of these food insecure households? What are the primary reasons Calaveras County residents are forced to seek

emergency food assistance?
What are the needs, barriers, and obstacles of Calaveras County residents that seek

emergency food assistance?

Survey findings addressing these questions are discussed in the next chapter and in the hunger report (Appendix A, pp. 42-82). Selected findings based on the cross-tabular analysis are also presented in the hunger report. These findings provided the information needed to refute the fallacies emergency food providers had identified as key issues in Calaveras County:

  1. Hunger is not a problem [in Calaveras County].
  2. The only hungry people in the county are bums and transients.
  3. The reason people are hungry is because they’re too lazy to work.
  4. Welfare payments are adequate to feed an individual or family.
  5. Hungry people don’t know how to manage their money.
  6. It doesn’t hurt to go to bed hungry once in awhile.

    (Contra Costa County Hunger Task Force, 1987, pp. 1-2)

The final document, “Calaveras County Hunger Report 2000: Voices of the People”, was compiled based on the results of the literature review, analysis of the survey results, and input from community stakeholders. This 40-page report (Appendix A, pp. 42-82) includes the recommendations for a community action plan (pp. 73-75) as well as a summary of the findings from the Client Survey, statistical data, and various other information relevant to hunger issues.

28