Community Needs Assessment
This chapter provides a brief introduction to needs assessment and its significance as a tool. Needs assessments are important tools for communities. They can have important impacts on policy-making decisions, provided they are carried out well and results are disseminated effectively. There are many different methodologies that can be used to conduct an assessment, and these will be discussed. It is also possible to conduct assessments on various subgroups of a community as opposed to the whole community. However, the importance of using an organized approach in any assessment cannot be overemphasized, regardless of whose needs are assessed and how they are assessed. This chapter will provide a background for organizing one’s approach to conducting an assessment, and it will provide useful resources for one interested in doing an assessment in the future.
Needs assessment itself has been given many definitions. The United Way of America (1982) defined it as “a systematic process of collection and analysis as inputs into resource allocation decisions with a view to discovering and identifying goods and services the community is lacking in relation to the generally accepted standards, and for which there exists some consensus as to the community’s responsibility for their provision.” McKillip (1987) defined it as a process of “ordering and prioritization of community needs.” Reviere et al. (1996) take a somewhat different approach in defining needs assessment. They approach needs assessment from the perspective of utilization of results/outcomes of the research. Their definition of needs assessment is “a systematic and ongoing process of providing usable and useful information about the needs of a target population- to those who can and will use it to make judgments about policy and programs.” It is a broader and more practical definition. Their definition is the one that will be used in this chapter, considering the goals of this text, since their definition has implications on policy-making decisions.
Reviere et al. (1996) note that needs assessment should be considered an applied social research, and that as such it should adhere to social research guidelines involving methodological design and data analysis. However, they note that methodologies used in needs assessments are at their beginning stages. Many previous studies have simply used methods that the researchers deemed sufficient for their goals. Deficiencies in methodology have arisen for various reasons. Assessments may not have been conducted by well-trained researchers; the researchers may have been under time pressure to obtain their results; and funding agencies may have put pressure, whether deliberate or otherwise, on researchers to publish results consistent with their views in order that they could influence policy-making decisions. The result has been that many researchers lack guidance in how to conduct a needs assessment, and that many assessments are conducted ineffectively. Lareau and Heumann (1982) determined that over three-fourths of a national sample of needs assessments of the elderly was “of such low quality… they could not provide meaningful input” into planning stages.
This does not mean that effective and helpful needs assessment have not and can not be conducted. However, it does speak to the need for adequately trained researchers as well as careful consideration of methodological issues in the design of an assessment. Planning is extremely important, and should be one of the first steps in an assessment; it should definitely be done before actual data collection. It should also be done throughout the assessment, and planning sessions should be incorporated into the process at regular intervals. Of course, changes will almost certainly need to be made along the way, and flexibility in the plans is important. However, as Reviere et al. (1996) note, “it is easier to make changes in a plan that is already in place than to start from scratch each time the unexpected happens.” Planning should involve consideration of funding for the project, goals and objectives of the project, definitions of individual research staff members’ roles and responsibilities, necessary supplies, methodological design issues, what data will be collected, how it will be collected and analyzed, and dissemination of the results of the assessment.
Further discussion of design issues, including choice of data to be collected and sampling techniques will follow in the next sections. However, a few brief notes should be made at this point about others of the above considerations. First of all, a clear statement of the goals and objectives of a project is extremely important in determining its success. Goals are abstract, more global statements of desired outcomes, whereas objectives are specific concrete statements about measurable outcomes. The goals and objectives should be explicitly stated early in the planning stages. Secondly, clear definitions of staff members’ roles and responsibilities helps to avoid confusion and disagreements among members of the research team. Lack of clear role definitions can ultimately slow the process down and result in difficulties in transitioning from one phase of an assessment to another. Finally, if the assessment is to have any impact on policy decisions, results need to be disseminated effectively and in a timely manner. Results may need to be disseminated to various audiences, using different manners of dissemination for each specific audience. Many needs assessments have been conducted without discussion of this step in the final write-up. It is the step of most interest to readers of the project, and it is the step that is most important in potentially bringing about beneficial changes that will affect study participants and the community.
The data to be collected in a needs assessment will be dictated by many factors. These factors include who is funding the project and their goals, qualifications and experience of the research team, the population being assessed, and the goals and objectives of the project. The key variables of interest to collect should be decided upon in the planning stage, so that the team can move on to determine sources of the necessary data and how to go about collecting them. Demographic variables are important in almost any assessment. Demographics to be obtained will be defined by one’s target population but those to consider include gender, age, race/ethnicity, education, marital status, and employment.
Data can be subdivided into two large categories, primary and secondary data. Primary data is data that is collected by the research team at the time of the assessment. Secondary data, on the other hand, is data that has already been collected, often by another organization or governmental agency. Examples of secondary data include the U.S. census, vital statistics registers, social indicator data, service utilization records, and data from mandatory health care/outcome reporting. Secondary data is often very useful and can save time and money; however, certain considerations should be taken into account when it is used in a project. It is necessary to consider ease of access to the original dataset (is a fee and/or permission required?); how the sampling was conducted in the original study; whether or not sufficient data is available on the population that the current team is interested in; and how the variables of interest were defined in the original dataset. Adjustments or modifications may need to be made in order to fit the use of secondary data into the research project. These adjustments themselves often cost time and money, and the research team should consider if and how such adjustments will affect the project.
Examples of primary data include surveys and interviews. The benefit of primary data is that the data collected can be determined specifically based on the goals and objectives of the project. The main drawback in using primary data is that it is often expensive. There is usually a longer time frame involved in using primary data as well. Decisions about what data will be collected, methodology involved in collection, analysis, and staff training are all factors that can increase the length of time necessary to complete an assessment. Using primary data offers greater control of the data collected, but it is beneficial for the team to weigh the costs involved in having greater control.
The above discussion may lead the reader to question whether primary or secondary data is more important. This is a question that should be answered on a case by case basis, as there are benefits and drawbacks of using either. The answer may depend on logistical issues such as cost, time requirements, experience and training of staff, and ease of access to data sources. However, it should be noted that one is not restricted to using one or the other, and that both primary and secondary data can be used in an assessment. In fact, resources permitting, this may often be the best approach. Reviere et al. (1996) comment that “there is no methodological reason to confine any needs assessment to exclusive use of either secondary or primary data… Careful and artful combination of secondary and primary sources… greatly strengthens a research design.”
Sampling Techniques, and Mode of Data Collection
This section will provide an overview of collecting primary data from the field, including sampling techniques and different modes of data collection. Since the most common method for primary data collection is a sample survey, it is important to understand the different methods of probability sampling. Lareau (1983) notes that one of the biggest problems with past needs assessments has been poor sample design. A good sample helps researchers to draw appropriate conclusions from the study, but a poorly designed sample weakens any conclusions drawn from the study.
Probability sampling, or sampling in which each individual in the target population has an equal chance of being selected for inclusion into the study, is commonly used in the social sciences. There are multiple types of probability sampling, and they are discussed in more depth by Cochran (1963). Fortunately, sampling techniques have not changed drastically in the past few decades. The first type of sampling is simple random sampling. Simple random sampling is a method of sampling where each individual in the population has an equal chance of being selected. Each individual is assigned a number, and then numbers are randomly selected until the required sample size is obtained. Cochran’s text provides a more detailed discussion of sample size, but a general rule of thumb for determining sample size is that a randomly-drawn sample of at least 30 individuals is sufficient. Simple random sampling is the most widely-used technique in probability sampling. Another type of sampling is called stratified random sampling. In this case, the population is broken down into two or more subpopulations of interest, the subpopulations being termed strata. For example, an assessment of preventive health screening in the elderly may break the population into strata by type of insurance that the individual has. After the strata have been created, a sample of each stratum is drawn, each stratum’s sample being independent of the others’. The benefit of stratified sampling is that it allows comparison of results based on the stratifying variable, and it may yield higher precision in estimates of characteristics of the whole population. A final type of sampling used in needs assessments is called cluster sampling. In this case, the first step of the technique involves numbering and randomly selecting units termed clusters. Examples of clusters include counties in a state, or city blocks. After the necessary number of clusters (generally at least 30) is selected, a simple random, systematic, or stratified random sample is taken from each of the clusters. Cluster sampling is often used in countries where there are no complete lists of people in a region, but where maps of the region permit it to be divided in units that can be randomly selected. The World Health Organization has used cluster sampling to estimate immunization coverage of their Expanded Programme on Immunization (EPI). An assessment of their results by Henderson (1982) concluded that the cluster survey technique performed satisfactorily.
The three main modes of data collection are mail surveys, telephone surveys, and in-person surveys. Each has its own advantages and disadvantages, and careful consideration should be given to which one or ones will be used in an assessment. In-person surveys are very expensive and they take time to conduct. However, they offer a more personal method of collecting data from respondents, and they allow for more direct examination of participants’ unmet needs. If respondents view services other than those previously defined as being necessary, this mode of data collection will be the one to catch them. Telephone surveys are less expensive than surveying individuals in person. When compared to mail surveys, they allow researcher to explore respondents’ attitudes more thoroughly. They are more expensive and require more time than mail surveys; however, response rates are usually higher with telephone surveys, and they often permit the collection of more detailed, quality information. Mail surveys are the least expensive and most basic of the modes of data collection. They are suited best to collecting straightforward facts and numbers. It is often necessary to send reminders or repeat survey mailings to participants to obtain the desired response rate. Reviere et al. (1996) note that researchers may draw a sample that is larger than the total number that is needed in order to ensure an adequate response rate. In using this approach, however, the research team should at least question whether there are any important differences between the groups who do respond versus those who do not. Such factors can affect analysis and how results of the research are used.
Nothing has been mentioned to this point about the use of new technology in data collection. Laptop and handheld computers have the potential to facilitate data collection and data entry, steps that traditionally have taken a significant amount of time. These instruments could be taken into the field, and either be used to administer surveys to participants by an interviewer, or they could be set up to allow participants to self-administer surveys. In either case, the data would be entered into the computer at the time of the survey, and could easily be transferred into a larger database at a research office. Another benefit of using computers to collect data is that skip patterns on survey instruments could be built into the program, which would facilitate data collection in terms of confusion and time. For example, in an assessment of access to health care among the elderly, suppose a questionnaire had a section about dialysis. The first question asked participants whether they need access to a dialysis unit. The following questions in the section then asked about ease of access to the unit. If a computer program was being used and the participant answered no to the question about needing dialysis, the program could be designed to skip the remaining questions in that section and move immediately on to the next section of questions. Obviously, this would make the survey easier and potentially faster to administer or take. The use of computers in data collection is likely to become more and more important as cost for these types of instruments goes down, as researchers become more familiar with their use, and as programs that allow more flexibility in survey design become more common.
The use of qualitative data in social sciences has recently been achieving more and more acceptance and popularity. It was not used commonly in the past because many researchers were not comfortable with the functionality and analysis of qualitative data. Qualitative data is data that is expressed in words as opposed to numbers, and it uses less structured and more open-ended methods of collecting data. However, the appropriate use of this type of data can add richness to an analysis. They allow for an issue or question to be probed more in-depth than with quantitative data. Different techniques are used both to collect and to analyze qualitative data. It is important to note that “qualitative methods do not usually fit with the assumptions or requirements of probability sampling. Unlike sample-survey methods, qualitative data collection and data analysis are quite labor-intensive, and usually focus on examining fewer cases in greater depth, rather than collecting data on large numbers of cases or respondents.” (Reviere et al. 1996)
Techniques in qualitative data collection include focus groups and intensive interviews. Qualitative data collection can also be incorporated into surveys with the use of open-ended questions. Of course, the comment above about probability sampling techniques does not apply to qualitative data used in this fashion. The discussion to follow will be restricted to focus groups and intensive interviews. Both instruments involve discussion of a specific topic using a less structured and more open-ended format. The difference between the two methods is that focus groups involve an interviewer facilitating discussion among a group of participants, whereas intensive interviews are usually one-on-one situations. In both, a general outline is followed by the interviewer so that issues related to the topic at hand are discussed in detail. Interviewers receive training so that they learn to achieve a balance between directing respondents to say certain things versus allowing them to turn to issues or topics that are not relevant to the subject matter at hand. The goal is to allow openness and flexibility in responses without straying too far from the topic being discussed.
Analysis of qualitative data is a separate subject, and will be discussed here only briefly. It is much more time-consuming than analysis of quantitative data. Miles and Huberman (1994) have described essentially three important components in the analysis of qualitative data: data reduction, data display, and conclusion drawing and verification. Data reduction involves transforming data in field notes or interview tapes so that they will be more manageable, and so that the elements most critical to the research question are selected. Data display involves organizing the reduced information into a form of display that will allow the data to be analyzed. Examples of displays include text, charts, and diagrams. Some analysis of the data occurs at this time as patterns and relationships between data become more obvious. The third broad component of qualitative data analysis is conclusion drawing and verification. This step involves figuring out what the data mean, how they help answer the research questions, and finally going back into the original data to verify conclusions that are drawn. Miles and Huberman (1994) comment that the three elements can often overlap in time. They note that “all three occur both during and after the data collection period and interact in a variety of nonlinear ways.” (Reviere et al. 1996)
As mentioned earlier, it is important to consider what will be done with the results of assessments once the phases of collection and analysis are completed. Results sitting on a shelf do no good, and can even jeopardize future assessments if the study participants feel that the research has had no beneficial effects. To have an effect on policies, programs, and delivery of services, the results need to be communicated effectively. Planning on how to present the results and to whom they should be presented should occur at all stages of the research project. Stakeholders in the assessment should be able to have input into presentations. At minimum, a final project report should be written and should include a brief summary as well as recommendations for change. Often results need to be presented to various audiences, and the research team is strongly encouraged to include the study participants as one of the audiences. The party who funded the project should be another audience. Presentation should be tailored to specific audiences, depending upon the needs, desires, and understanding of each. The importance of disseminating results effectively using these guidelines cannot be stressed enough. A needs assessment will have achieved nothing if the results are not communicated effectively.
Needs assessments are important tools for determining if and how changes should be made, especially in the provision of health care. The current state of the field is exciting because many opportunities and challenges are presented. Understanding of methodologies involved is essential in conducting an effective assessment and in using results to advocate for change. Many assessments in the past have used methods of data collection that are not up to par with methodologies in other fields. It is important to train researchers in the thoughtful and appropriate use of diverse methods used in collecting data for an assessment. The use of qualitative data in needs assessments is achieving more and more acceptance as a method, and it has the potential to broaden and enrich the field as a whole as well as assessments that use analysis of qualitative data. Technologic advances, most importantly advances in computer technology, may be the most important recent changes in the field. They will allow for greater ease in data collection and potentially more rapid turnover of results that can effect policy changes. It is hoped that these factors will encourage more researchers to become involved in needs assessment. Conducting effective needs assessments will allow effective advocacy for positive community changes.
1. Acosta O and Toro PA. Let’s Ask the Homeless People Themselves: A Needs Assessment Based on a Probability Sample of Adults. American Journal of Community Psychology 2000;28(3):343-66.
2. Cheung KM. Needs Assessment Experience Among Area Agencies on Aging. Journal of Gerontological Social Work 1993;19(3/4):77-93.
3. Cochran WG. Sampling Techniques. 2nd ed. New York (NY): John Wiley & Sons, Inc.; 1963.
4. Ensign J and Gittelsohn J. Health and access to care: perspectives of homeless youth in Baltimore City, U.S.A. Soc. Sci. Med. 1998;47(12):2087-99.
5. Ervin AE. Trying the Impossible: Relatively “Rapid” Methods in a City-Wide Needs Assessment. Human Organization 1997;56(4):379-87.
6. Harlow KS. Research and the Politics of Decision Making: Planning Services for Elders. The Journal of Applied Gerontology 1992;11(1):22-37.
7. Henderson RH and Sundaresan T. Cluster sampling to assess immunization coverage: a review of experience with a simplified sampling method. Bulletin of the World Health Organization 1982;60(2):253-60.
8. Lareau L. Needs assessment of the elderly: Conclusions and methodological approaches. The Gerontologist 1983;23(5):518-26.
9. Lareau L and Heumann L. The inadequacy of needs assessments of the elderly. The Gerontologist 1982;22(3):324-30.
10. McKillip J. Need analysis: Tools for the human services and education. Newbury Park (CA): Sage; 1987.
11. Miles MB and Huberman AM. Qualitative data analysis: an expanded sourcebook. 2nd ed. Thousand Oaks (CA): Sage; 1994.
12. Reviere R, Berkowitz S, Carter CC, Fergusen CG, editors. Needs Assessment- A Creative and Practical Approach for Social Scientists. Washington, DC: Taylor & Francis; 1996.
13. United Way of America. Needs assessment: The state of the art. Alexandria (VA): United Way of America 1982.
14. Williams RL and Yanoshik K. Can you do a community needs assessment without talking to the community? Journal of Community Health 2001;26(4):233-47.