Gq-logo2.gif (1077 bytes)


Title_da.gif (5003 bytes)

 
overview.GIF (371 bytes)
How to use this manual
What is formative research?
What do you want to know?
   
before.GIF (511 bytes)
2a-handle.gif (559 bytes)
2b-ensuring.gif (485 bytes)
2c-getting.gif (344 bytes)
 
datacoll.GIF (610 bytes)
3a-archival.gif (345 bytes)
3b-field.gif (528 bytes)
3c-focus.gif (344 bytes)
3d-depth.gif (417 bytes)
triangle.gif (460 bytes) 3d-surveys.gif (272 bytes)
3f-innov.gif (375 bytes)
 
resources.GIF (370 bytes)
Search
pdf.GIF (352 bytes)
publish.GIF (408 bytes)

Surveys
, continued
< backnext.gif (239 bytes)


Using outside agencies

There are basically three ways to go for inputting, cleaning and analyzing data: do it by yourself; use technical assistance to help as you do it; and hire someone else to do it. Any outside agency should be involved from the beginning of the project, from design or adaptation of the survey to pilot testing to implementation.

If you conduct data entry and analysis by yourself, there are several books and manuals that can help you in the process (see Appendix 5 of the Resources section). Agencies should consider supporting staff training in order to have someone on board who can handle data analysis. Classes can be found at the local college, at conferences or through technical assistance agencies (see Appendix 6 of the Resources section).

Case Study
  In San Diego, the technical assistance team conducts a community-wide evaluation every year focusing on different target audiences. Each agency fills out surveys among their clients and submits them to the TA team. Tom’s agency didn’t hear anything back after the survey in 1996, and he plans to make sure they get a report back from the 1998 survey on injection drug users. “I won’t have the data. It’s being compiled by consultants, so they’ll have it all. But we’ll ask for copies of what the findings are. And we’ll ask for them to split out the data according to where the surveys were collected. Because our injection drug use population is very different from say, South Bay, where you get a lot of Hispanics using heroin. We have a lot of white males shooting crystal out here, so it’s going to be a real different psychographic makeup of these people. They could do a separate analysis for all six counties and it would be helpful.”

Park.jpg (11535 bytes)Park in El Cajon where injection drug users hang out.

 

 

Agencies can also use technical assistance resources to problem-solve or give advice during the process. If the survey is not overly complicated and a staff member has experience in or is willing to learn data analysis, technical assistance can provide supervision and help on specific rough spots (see Appendix 6 of the Resources section).

Agencies can also hire an outside researcher or negotiate for a pro-bono donation of time. Researchers in many universities are eager to work with community service providers. The American Psychological Association (APA), Office on AIDS has a special program that links researchers with agencies (see Appendix 6 of the Resources section for a further description of this program). There are also many private organizations that provide research and data analysis (also listed in Appendix 6).

Some agencies use graduate students in Public Health or other social or behavioral sciences as interns to input and analyze data collected. Most Public Health candidates are required to work at an agency as part of their education. Graduate students need to be trained and supervised when working with community-based agencies.

If you are using an outside agency to do the data analysis, clarity and communication are essential.

•    Make sure the jobs are clearly defined before starting. Who will input the data? Who will code it? Will there be a written final report? How detailed will it be?

•    Make your data needs known. If you are interested in ethnic differences in respondents, or comments on what didn’t work in your program, let them know ahead of time. Most data can be analyzed any number of ways, and you need to make sure that the outside agency and your own agency agree on what you’re looking for.

•    Discuss preliminary findings together. This step is most often overlooked. Often, researchers and service providers can interpret the same findings in different ways. If you both agree on what the preliminary data suggests, then further, more in-depth analysis can address the same questions.

Further reading on surveys can be found in Appendix 5 of the Resources section.

< backnext.gif (239 bytes)

 

Good Questions, Better Answers --  � 1998 California Department of Health Services and Northern California Grantmakers AIDS Task Force  -- http://www.goodquestions.com