Runkle and McGrath (1972) have advised that the research
strategy chosen for a needs assessment should be the most appropriate for
the purpose and circumstances. Since there is no one ideal strategy it
was decided to use a variety of assessment procedures in this situation
to tap as many consumer and provider sources as possible. The needs assessment
was conducted in three steps. A survey to services interrelated with mental
health was mailed first, since those in related helping professions presumably
would be most knowledgeable about needs and service delivery. The second
step was to carry out a county-wide survey of residents. The final phase
was to offer community forums. Since each step involved somewhat different
procedures each is presented in turn.
To gain perspective before conducting a needs assessment it is useful to review related research on the geographic area to be studied. Such a review can save time aid in constructing questions and provide a basis for comparison. County extension services school systems and other community agencies often are excellent sources of such data.
In March 1977 a mental health quality-of-life questionnaire was sent to 205 agencies, physicians, clergy, school officials and law enforcement personnel in the county seat and in the two other towns large enough to have a Chamber of Commerce. Chambers of Commerce have lists of community leaders and organizations. Respondents were requested to rate how adequately 21 service needs were being currently met on a 5-point scale which ranged from always (100%) to not at all (0%). Don t know was a sixth choice. These needs ranged from outpatient mental health care to recreational facilities. Respondents also were asked to list the three greatest needs for each of the service areas to supply any needs not included on the instrument to rank their top three needs priorities and to supply any additional comments. To enhance the response rate a cover letter stamped return envelope and when necessary two phone calls offering to answer questions about the instrument were employed. The questionnaire which was unsigned and visibly coded was designed to elicit both evaluations and recommendations.
The method of mail surveys and follow-up phone calls was effective in achieving a 41% return rate. The response rate from the county seat was 52% from the next largest town 37% and from the third largest town 9%.
Means, omitting the don't know responses, were computed on all the services for each town separately. This process provided insight into needs of each town. Another helpful analysis was dividing respondents from a given town into categories such as agencies, clergy, physicians, etc. This division into categories highlighted the groups which might be inclined to support programs directed at particular needs.
In general, it appeared that many of the 21 programs needed additional publicity. For example 32% of the responses from the county seat were in the don't know category. From the next largest town located 18 miles from Muscatine 65% of the responses were don't know. A 59% response rate in this category was obtained from the third largest town located 10 miles distant from the county seat. The principle appears to be that the farther key personnel are from the population center the less informed they are about services. Additionally better known services such as recreational facilities, services to the elderly and services to minority groups tended to be perceived as more adequately meeting specific needs that the relatively unknown specialized programs such as aftercare.
The questions which requested the listing of additional needs, the ranking of the three greatest needs as to priorities, or the writing of the three greatest needs for each of the 21 services were not as productive as hoped. The last item above was the least fruitful and consequently was not included in the community survey.
One additional benefit of surveying the interrelated services first and having the write-in questions was that it became clear that some questions required revision. For example day care and aftercare had to be better defined for these terms were new even to people in mental health related areas. Additionally on the basis of the respondents comments other services were added to the instrument for the next step. These Included such Items as coordination among agencies, services for low-income families, mental health consultation, and services for school dropouts.
In July 1977 the proposed questionnaire and cover letter
for the community survey were sent to the Directors of the Muscatine Community
Health Association which directs planning in health areas. They supplied
many helpful suggestions and the final questionnaire was revised accordingly.
Contacting a group like this provides a more general review of the construction
of the questionnaire and gives it a trial run with a diverse sample of
The instrument for the county-wide survey now contained 33 service needs to be rated in the same fashion as in the first survey. Additional needs also could be listed.
To assess whether previous public education efforts by the mental health staff were successful community residents also were asked if they knew of the existence and the location of the mental health center before they received the questionnaire. Residents who had heard of the center were requested to mark the various means by which they had learned of it (eg. Talks given by center staff, media presentations, word of mouth, other). Those who heard talks also were asked to rate them.
Selecting a random, representative sample for a rural population is difficult. Baumel and Hobbs (1964) delineate methods for obtaining a random sample of households both from county plat books which include unincorporated areas, and from city directories. . A fiscal report from the county auditor's office offers statistics on cities, towns, and townships from which the correct sample number for each segment of the population to be surveyed can be drawn.
An adequate sample of households for a county of Muscatine's size (37,181) is 400 (Baumel & Hobbs 1964). Two hundred forty-one households were drawn from a city directory. The 101 selected from the plat book were representative of a geographical area, even though the plat book lists the name of the property owner who may or may not be living there. The biggest problem was to obtain a random sample of the 58 households not in the plat book and in towns without a city directory.
Public officials of each of these towns were contacted first and asked for names of residents. No replies were received. Then, the power companies were contacted. Since their lists had different boundaries, they could not be used. Although excluding some people with lower socioeconomic status, the phone book for each town finally had to be used to obtain the most random sample possible under existing conditions.
The questionnaire was mailed in the same manner as the first instrument. Follow-up phone calls were made. Those not returning the survey were sent another one within a few weeks after the first mailing. Those on the follow-up mailing were again contacted by phone. Only 20 questionnaires were returned by the post office as undeliverable. No substitutes were allowed in the Baumel and Hobbs (1964) sampling method.
Geographically, the survey had been mailed on a proportionate basis. The respondent sample was divided into four areas according to townships. The chi-square analysis indicated that a representative, random sample of the county as to place of residence had been obtained, x2(3)=.20, ns.
The proportion of "don't know" responses was again higher from people outside the city and township in which the county seat was located. T-tests were run on each of the 33 services evaluated, excluding the "don't know" responses. In two instances, there was a greater need (p<.05) perceived outside the county seat. An analysis such as this highlights targets for increased public education and/or program development.
In the county wide survey, listing additional needs "not covered on the instrument" was even less productive than the item was on the interrelated services questionnaire. Most of the "other needs" already were listed. Ranking needs according to priorities drew an even smaller response.
More useful were the questions evaluating the center's community education efforts in the 77% of the entire sample knew of the center and 57% knew where it was located before they received the questionnaire. Again those community residents living in the county seat were more knowledgeable about the center's existence (t = 4.91, p < .01) and location (t = 6.11 p < .01) than those in other areas.
The center's community education program had used talks by staff, newspaper articles and announcements, radio programs and announcements, brochures, coffees for other agencies, etc. More than 40% more respondents indicated that they had heard of the center via newspapers than from the next largest category word of mouth. Newspaper coverage thus became a priority focus in community education.
Those who had heard talks by center staff evaluated them.
Results were that 20% found the programs very informative and another 50%
thought them informative. Such feedback was reinforcing to the Consultation
and Education efforts especially since no one marked that they were dissatisfied
or extremely dissatisfied with the talks.
Community forums were arranged for the county seat and
for the next two largest towns. An article describing the purpose of these
meetings appeared in the newspaper published in the county seat and public
announcements were placed in the newspapers of each of the three towns.
Center board members who resided in each of these locations made arrangements
for their respective towns and suggested who should receive individual
announcement of the forums. In spite of these efforts fewer than 10 people
came to the three forums combined. Other than giving everyone a chance
to express his or her views and having no one show up primarily to vent
complaints the community forum was the least successful assessment procedure
utilized. Basically no new information emerged from them.
The purpose of a needs assessment is to ascertain priorities for mental health planning. The results from each of the assessment techniques were scrutinized for common trends. In the present study rape prevention and control and a transitional half-way house emerged as the greatest perceived needs in the majority of groups and subgroups studied. The rape prevention and control program is now in place In the county and a transitional half-way house for the developmentally disabled also is operating. The contribution of the needs assessment was to help focus attention and to provide additional data regarding these needs.
The needs assessment was distributed to community leaders and funding agencies (i.e. county supervisors United Way). Center board members volunteered to distribute it to key persons in all of the county towns. Various board members also put up posters which listed the center's programs in places where people congregate (e.g. Banks, general stores, post offices, and bars). This was one effort to have continuous visibility in each community.
The needs assessment also aided the center's programming. For example it was learned that aftercare was a program that most surveyed groups felt was needed. However the center had established an aftercare program in October 1975 but had not widely publicized it. Action steps were:
1. The existing aftercare program was described in the needs assessment report.
2. A center newsletter was started. One of the feature articles in each issue was to be a description of a specific center program. Aftercare was the first one selected. By obtaining a nonprofit organization stamp it was possible to mail more than 500 newsletters to various community leaders at very little cost.
3. Two staff members went to the local radio station to give two interviews: one on the available mental health services and the other on the aftercare program.
4. During a 1-week period the newspaper in the county seat featured a series on mental health including an article on the center and one on the mental health institute which provides aftercare services.
5. Articles on aftercare also were sent to other county newspapers.
Thus data from the needs assessment influenced community planning and to a larger extent the mental health center's program. The assessment described here sought both evaluations and recommendations from community sources. Despite procedural problems the survey of providers interrelated with mental health and the survey of county residents were most productive in providing input. Both also produced good response rates—41% and 58% respectively. Two principles emerge overall from these data: (a) that the closer community leaders, providers, and residents are to services geographically the more they tend to be informed about them; and (b) that services which receive wider usage such as recreational facilities transportation or school programs are those which respondents appear to be the most informed about and more inclined to evaluate.
Needs assessment results will vary from area to
area. The contents of this paper may aid those concerned with the pragmatic
steps involved in conducting and utilizing the results of needs assessments.
With the present Zeitgeist stressing accountability needs assessments and
other evaluative procedures are more necessary than ever. Concern also
is being expressed (Schulberg & Perloff 1979) about how best to train
evaluators for human service delivery programs. As a professional psychologist
whose graduate training did not include these types of evaluations and
as one who was faced with carrying them out I believe that training in
both service delivery and in measurement techniques is essential for developing
evaluations that are pertinent professional and politically sensitive.
Baumel C.P. & Hobbs D.J. The community survey: Its use in development and action programs. Ames: iowa State University Cooperative Extension Service 1964.
Runkel P.J. & McGrath J.E. Research on human behavior:
A systematic guide lo method.
New York: Holt Rinehart & Winston 1972.
Schulberg H.C. & Perloff R. Academic training of human
service delivery program evaluators.
American Psychologist, 1979, 34, 247-254.
Zautra A. & Simons L.S. An assessment of a community
s mental health needs. American
Journal of Community Psychology, 1978, 6, 35-362.
(Original journal pages 46-50)