Dear list serv,

Several questions dealt with assessing suitability  of pictures - are they 
understood and accepted by the intended audience?   Focus groups were suggested, 
but caution was advise to prevent making a decision  based on "the loudest 
voice" in the focus group.  Also, what people say  they want may not in reality 
may not turn out to be what they  want.   Here are some practical suggestions 
to respond to these  issues.
 
Focus Groups versus one-on-one interviews:  For community issues, for  
example, environmental problems or general health  promotion behaviors, focus groups 
are  useful and  effective.  However, for specific clinical instructions or 
sensitive issues  such as AIDS or STDs, the one on one interviews seem to be 
more productive  because the interviewer can probe answers, especially for 
patients with low  literacy skills. Another time to use one- on- one is to  verify 
the  discrepancies that may arise in focus groups.  In these instances,  
verification with a sample of the intended audience is most helpful.   For example, 
we verified diet instructions for kidney dialysis patients  directly in the 
clinic while the patients were undergoing treatment.
 
The key question in testing visuals is "will the visuals (art  work/graphics) 
help or confuse the key points of the message?"  Three areas  to cover are:  
(1) Let's look at the cover.  What catches your  eye?"  (2)  What do these 
pictures tell you?  (3)What do you  think of the color?  It is important that the 
visuals portray familiar  scenes or objects; this is why "modernistic" art is 
less likely to be  acceptable, it is not familiar.  People need to be able to 
identify with  the visual presented and the action recommended.  Types of 
questions to ask  are included in the chapter on "Learner Verification & 
Revision" in Teaching  Patients with Low Literacy Skills", Doak, Doak, & Root, p. 174. 
 This  is available free, on-line at the Harvard School of Pub Health web 
site.
 
Another issue is that of cost and time needed to make such field tests,  
especially the one to one assessments. The answer:  Not very many - maybe 5  or 
10. For example: While serving as consultants to their "Eat five a day"  program 
for seven different ethnic groups of Amricans,  the NIH  statisticians 
determined that 30 subjects (three different groups in one to one  testing) was 
suffiecient.   In evaluating the verbatim comments to the  test questions of those 
who were presented with the draft instructions, we found  that if anything 
was wrong it showed up in the first ten subject  responses.  
 
Does this help?
 
best wishes,
 
Len and Ceci