Seeking clarity on outcome terminology? Want to practice writing a survey? Need to devise ways to track and report outcomes? If so, sign up for Measuring Program Outcomes, a thorough, hands-on workshop on outcome-based evaluation. The Fall sequence of sessions will meet September 13, 20, and 27, and October 4 (Mondays), 1-4:30 pm in the Greater Kalamazoo United Way Board Room (709 S. Westnedge Avenue). Cost is $15/person; any community member welcome. Register by September 3, 1999. For a registration form, call Elaine Griffin at 616/343-2524.

Writing good questions elicits good data! We need good data to guide our efforts at program improvement. Whether you use a printed, self-administered questionnaire or an interviewer who reads from a script, the questions must ask for information clearly. Response rates and accuracy will be highest if we avoid survey questions that make respondents feel threatened or frustrated. Pamela Narins of SPSS, a social science software company, offers suggestions for effective survey questions.

1. Remember your survey's purpose: Whatever prompted you to spend your time and money on a survey should be the focus of every question. If you cannot directly link a question to the survey's purpose, take Narin's advice: "When in doubt, throw it out."
   
   
2. Keep your questions simple: Use simple, declarative sentences. If you find yourself writing long, compound sentences, break them into component parts. Make each part a separate question. Otherwise, respondents have to keep much information in their heads at once, and may respond in odd ways.
   
   
3. Include only one topic per question: "Please rate your satisfaction with the amount and kind of care you received in Program X." If the respondent answers "Very Satisfied," is the person happy with the amount of care, the type, or both? Avoid "double-barreled" questions.
   
   
4. Be specific: "Did you learn something from Program Y?" is fine as a general question, but would not provide enough information for program refinement. For that, one needs more specific questions on particular skills or knowledge acquired in Program Y.
   
   
5. Be clear: "What time do you eat dinner?" may be interpreted differently by people in urban and rural areas. Less likely to be misinterpreted would be the question, "What time do you eat your evening meal?"
   
   
6. Avoid leading questions: Try not to give respondents the impression that a question has a "right" answer. "Most experts believe that youth need structured after-school activities. Do you agree?" Such leading questions may prejudice the results of the survey.
   
   
7. Consider alternate ways to ask sensitive questions: Sensitive topics include drug or alcohol consumption, sexual habits, income and age. Instead of asking exactly how much money respondents make, you may offer a set of income ranges and have respondents check the range within which their income falls. Instead of asking someone's age, you can ask what year they were born.
   
   
8. Make sure the respondent has enough information: When questions require background information, you need to check the respondent's awareness before you ask for a response to the issue. First ask a screening question describing the topic to see if the respondent knows about it, then ask a follow-up question addressing the respondent's attitude about the topic.
   
   
9. Keep open-ended questions to a minimum: Open-ended questions can elicit valuable information, but too many can tire respondents and pose a challenge when you try to aggregate and classify responses.
   
   
10. Response options need to be mutually exclusive and exhaustive: If you provide response options, be sure the options are mutually exclusive. If the choices for "number of children" are "1-2," "2-3" and "3 or more," which option does the parent of two children select? Be sure response options cover every possibility. If you are unable to provide a complete list of options, offer "Other" as a choice.
    
    
11. Define temporal terms: "Always," "Sometimes" and "Never" mean different things to different respondents. Specify what you mean by each term so that all respondents are using the same frame of reference. "Do you read the Program Newsletter regularly, that is, at least three out of every four issues?"
    
    

"Our Program Funding Request form used to ask for narrative responses to a number of questions, as well as for a program logic model and outcome results. We decided that an expanded logic model could serve as the entire Funding Request. As we took the old questions and fit them into categories of the expanded logic model, we realized that our questions about community collaboration and staff demographics presumed that these practices yield the best outcomes. The proposed revision of the Fund Request form omits those questions, which would leave it to programs to demonstrate which practices produce the best results."

--Janice Maatman, Greater Kalamazoo United Way

• Everyone’s  responsibility
• Continually asking good questions, getting answers, and taking action based on those answers
• Integrated into the day-to-day operations of the organization
• A developmental process
• Collaborative and dependent on information sharing
• Time well spent
• Going to ensure the organization’s health and viability in the long run in a changing environment