The following comments were drafted in August, 1998, by Kendall Repass, formerly with Peace Corps/Senegal and presently at Columbia University.

The strengths of the Scenarios from the Sahel questionnaire are as follows:

The questionnaire_s focus_information sources on AIDS for African youth in the Sahel_is unique and very much needed. Very few studies have examined this area in detail. One of the primary means of preventing AIDS is through AIDS education. This is especially true in the Third World where current drug therapy methods are prohibitively expensive. AIDS prevention workers therefore must know the most effective combinations of media, institutions, and persons through which AIDS prevention information should be channeled.

The questionnaire is short, limited to two sides of a single page, and is attached to the contest rules brochure. The questionnaire is therefore unlikely to be lost and doesn_t present an overwhelming amount of questions for contestants to respond to.

Most questions are clearly marked or numbered with clear instructions such as _if yes_, _if no_, and _mark all items that apply_.

With the exception of question 3.8, all questions with an _other_ response include a checkbox for _other method_, or _other source_ as well as a space to write in what the other response actually is. This anticipates what should be done in the data-entry process where each respondent should have a possible code of _other_ without specification. This "without specification" category is used when the researcher wishes to know only the number of persons having chosen _other_ as a response. Another variable is reserved for the actual comments and can be analyzed accordingly by any researcher who wants to know what the specific _other" categories are.

Questions 2.2 _I_ve heard about the Scenarios contest from_, and 2.3 _Where have you found the contest guidelines?_ serve as a built-in evaluation of Scenarios advertising and Scenarios questionnaire distribution. This is an excellent idea and should be continued in future Scenarios surveys.

Questions 3.3 _Have you sought to complement or add to the information you have received?_ and 3.8 _Have you talked about AIDS with__ distinguish between active information seekers and passive information receivers. Active and passive classes of respondents can be compared with other questions such as _I know someone living with HIV or AIDS..._ in order to determine if any relationship exists between the questions. In other words, does the fact that youth know someone with AIDS inspire them to become an active information seeker? Do active information seekers tend to be older, in school, or out of school, etc.? If question 3.8 is to serve this purpose, it would be better to ask two active/passive questions as follows: _Has someone else initiated a conversation with you on AIDS? If yes, who_?" followed by, _Have you yourself initiated a conversation with others on AIDS? If yes, with whom_?"

The last question, _I know someone living with HIV or AIDS (yes, no, don_t know)_, in addition to finding out how many youth actually know someone with HIV or AIDS, serves as an excellent knowledge question. The question minimizes cheating and doesn_t impose any ideas upon the respondent.

The questions in section three of the Scenarios from the Sahel survey may at first seem redundant. This is not at all the case, however. The series of questions addresses the many nuances inherent to the subject of information sources, covering the essentials of who, what, how, how much, and how useful. In fact, as the next section will show, even more questions of this nature should have been included.

The weaknesses of the Scenarios from the Sahel questionnaire are as follows:

There are no instructions preceding the questionnaire. This is a serious oversight considering that a good number of participants had never filled out a survey. Even in industrialized countries, a set of instructions is provided for most questionnaires as well as individual sections of questionnaires. The instructions should explain what a checkbox is (if they are to be used). The instructions should explain how respondents should mark their answers (circle, check, write on a blank line_) and that for some questions only one answer is allowed and for others one or more can be chosen. Examples of correctly filled-out sample questions should be provided. In closing, the guidelines should mention that further instructions will be provided on a section-by-section, question-by-question basis when necessary. The respondents should be reminded to answer all questions to the best of their ability and that the number of _incorrect_ of _correct_ answers will not influence their chances of winning.

The questions in section one, sections of question 2.1 (_the number of members in your team_, _the average age of your team__) and the _if yes, where_ question after question 3.3 are not numbered. All questions in surveys should be numbered, even if they are _part_ of the preceding question. This allows for easier data entry and coding and helps guide the respondent through the survey.
The existing survey is comprised of mostly practice and attitude questions. Only one knowledge-based question is asked_question 3.9. In addition to the existing attitude questions, the following attitude questions should be considered:

C _In your opinion, is the dissemination of information on HIV/AIDS an effective tool in the fight against AIDS? (yes, no, don_t know) If yes, why? If no, why?_.

C _Can this information actually cause you to change your sexual behavior? (yes, no, don_t know)_.

In the report _Observations and Recommendations of the Selection Teams_ [an internal Scenarios from the Sahel document drafted after the selection process was completed], a number of common participant errors were noticed, including moralizing, over-emphasis on migration, and misunderstandings of non-symptomatic seropositivity. The next Scenarios survey should address these knowledge issues. Questions such as the following could help remedy this situation:

C Can you tell if someone is HIV+? (yes, no, don_t know).

C How long does it take for someone who is HIV+ to become seriously ill?

C Who gets AIDS?

Since very few respondents wrote about STD's, knowledge questions on STD's should also be included.

Apparently, the question _I spent _____ years in school_ while ideally worded for coding purposes, was widely misunderstood or difficult for youth to answer. Youth probably think in terms of their current class or last class attended and not the number of years spent in school. In the Peace Corps, Plan International, and Burkina Faso sample, less than a quarter of participants answered this question. This missing data is highly unfortunate because respondents_ education level is an important variable in any study. The next Scenarios survey can solve this shortcoming by incorporating the following series of questions:

C Are you a student? (yes, no)

C

C If you are a student, what is the name of your school?

C If you are a student, what class are you currently in?

C If you are not a student, what was the last class you successfully completed before leaving school?

C If you are not a student, do you have a job? (Yes, No)

C If you are not a student and you have a job, what is your current occupation?

Fortunately, the vast majority of student participants answered the question _The name of my class is_. However, with greater amounts of non-student participants expected in the coming Scenarios project, the above series of questions will be essential in order to assure that education level is known for all respondents, not just students.

In addition to knowing participants_ education levels and occupations, respondents_ parents_ education level and occupation should also be determined. The effect a parent_s background has on his or her children is well documented in numerous social studies. In the health field, for example, it has been observed that children born to educated mothers have lower infant mortality rates.

There is a great deal of potential misinterpretation arising from the results of the question _If I want to watch television, I can do it (at my house, near my house_). According to the results, 44% of respondents in small towns and villages have a television in their home. Is this possible? Probably not. More likely, respondents marked their home address (say in a village) and responded to the above question in light of their current situation where they are attending school in a town or city. These two occurrences lead to a false interpretation of the actual situation. Furthermore, urban respondents are most likely confused with rural respondents and vice-versa. Many African youth are forced to leave their villages to attend institutions of higher learning located in towns and cities. Many youth leave large urban areas for smaller towns or even villages as a result of their or their parents_ job posting, especially if they or their parents are government workers. Since there may be differences between practices, knowledge, and attitudes between urban and rural populations, how can the researcher know who is who? In order to avoid these problems in the future, the following series of questions should be asked instead:

C For the next series of questions, please keep in mind the following definitions:

- Villages are defined as having less than 10,000 inhabitants.
- Towns are defined as having between 10,000 and 50,000 inhabitants.
- Cities are defined as having over 50,000 inhabitants.

C I have spent most of my life in_____ choose only one: (a village, a town or a city, in both a village and a town or city).

C Currently, how many months in the year do you live in town or in the city? If you haven_t lived in a town or city recently, mark 0.

C If you currently live 1 month or more in town or in the city, why are you there? choose only one (I live there, to visit family, to work, to attend school, other please specify_)

C When you are in town or in the city, where do you watch television? choose only one (where I live, near where I live, far away from where I live, I don_t watch television, I_m never in town or in the city)

C When you are in town or in the city, how often do you watch television? choose only one (every day, most every day, only a few days of the week, I rarely watch television, I don_t watch television, I_m never in town or in the city)

C Currently, how many months in the year do you live in the village? If you haven_t lived in a village recently, mark 0.

C If you currently live 1 month or more in the village, why are you there? choose only one (I live there, to visit family, to work, to attend school, other please specify_)

C When you are in the village, where do you watch television? choose only one (where I live, near where I live, far away from where I live, I don_t watch television, I_m never in the village)

C When you are in the village, how often do you watch television? choose only one (every day, most every day, only a few days of the week, I rarely watch television, I don_t watch television, I_m never in the village)

Taken together, these questions help determine the degree of the urban or rural status of a respondent and where and how often they watch television in rural and urban settings. These same questions also provide information on the mobility of youth, itself an important variable in AIDS research.
Questions on ethnicity and religion should be added to the survey. However, obtaining this information is more complicated than it seems. Some Africans who have moved to urban areas, especially the children of such urban immigrants, are culturally similar to an ethnic group different from their own, usually the predominate ethnic group in that location. In order to better discern such nuances, two questions can be asked:

C What ethnic group do you belong to by blood? (If you come from a mixed background, that is your parents belong to different ethnic groups, please mention both of these groups in your response)

C What ethnic group do you most resemble in your language, mannerisms, and culture? (Your response may or may not differ from the answer you gave above)

Similarly, nuances can be determined for religion as well:

C What is your religion? (Catholic, Muslim, Jewish, Protestant, Other please specify_)

C Do you consider yourself to be more religious than most people your age, as religious as most people your age, or less religious than most people your age?

In other words there are varying degrees of piousness among people of the same religion or across religions, and the degree of religiousness may be linked to differences in knowledge, attitudes, and practices concerning AIDS.

More information is needed on the groups participating in Scenarios. In addition to the questions already asked, it would be helpful to also know the number of literate members in the group, and group breakdowns according to their religions, occupations, education levels, and parents_ educational and occupational backgrounds.

The fact that some questionnaires are filled out by individuals and others by groups complicates the analysis of the surveys. The majority of questions address the person filling out the questionnaire. This is the case even if the number of questions pertaining to groups is increased as outlined above. The answers given by the group leader may or may not be representative of others in the group. It would be far better to require each group member to fill out a separate survey. Illiterate participants would be helped in filling out the survey by their literate friends. All respondents would continue to be asked if they are working in a group or individually. Those working in a group would be asked to provide a unique name for their group, and the name of the group leader. It would then be up to the group leader to fill out the section on group information including the number of males and females in the group, their average age, the number who are literate, etc. as above. Clear instructions would be provided so that all other group members would leave this section blank in order to avoid duplication and the analysis problems such duplication would cause. This group section would be best placed at the end of the survey, while the participating as an individual or as a team, group name, and group leader_s name questions would be placed at or near the beginning of the questionnaire.

Question 3.7 should be reworded in order to encourage more specific answers from respondents. The revised question may be written as follows: What exactly would you like to learn more about concerning HIV and AIDS?

Question 3.6 is confusing. Are we talking about a person or a place? The question should be divided into the following two questions: Where would you like to learn more about AIDS? (response must be a source, institution, or place), and Who would you like to learn more about AIDS from? (response must be a person)

The two above-mentioned questions along with questions 3.3 _Have you sought to increase or add to the information you have received_ and 3.5 _Where have you received the most useful information on AIDS?_ should all be followed by the question Why? Respondents may mention that they have sought to learn more about AIDS because they are afraid, are at high risk, or have had an STD. Respondents may mention that they prefer to learn more about AIDS from doctors because they are considered to be experts. The answers to these why questions could be quite illuminating and would eliminate a lot of guesswork on the part of the researcher.

Question 3.1 should be reworded as follows: From whom have you received information on AIDS, not where.

Father, mother, brother, and sister should all be listed as separate answers. In addition, the categories of friends, health agent, and teacher should be broken down into the following possible responses: male friends, female friends, health agent, sage femme, male teacher, female teacher. It is likely that youth are most comfortable speaking about AIDS with individuals of the same sex. Whether this is true or not is important for researchers to know. Given the current wording of these responses, this phenomenon cannot be verified.

Conference and "causerie" may be one and the same in the eyes of many youth. If so, only the most widely used of these terms should be used. Other terms should be reviewed as well during pre-testing.

Given the fact that Scenarios is to be repeated in many countries throughout the world, the same questionnaire should be used in order to enable cross-country comparisons. Each question and response would of course be _translated_ not only into different languages, but also into local terms in order to assure questions and their responses are understood in each country or culture. It goes without saying that the above commentary as well as the recommendations concerning the strengths and weaknesses of the current Scenarios survey should be incorporated into future questionnaire design.