See More: Articles

Needs Assessments: How to do Them, How They Help

Gelman, Rosenberg & Freedman CPAs is a member of CPAmerica International, an association of CPA and consulting firms that provides industry knowledge including insightful articles, to help member firms serve clients and other individuals and organizations.

You may have a good understanding of your clients, but doing a needs assessment on a regular basis can hone your programs and make the case for funding.

The results may confirm your beliefs or surprise you by revealing some unexpected information. The needs assessment tool can also be used to gather feedback about your services and suggestions for improvements.

In some cases, you may be able to use results for advocacy if they shed light on important issues. For example, a small business credit survey showing lack of access to loans was shared with federal policy leaders. Subsequently, the SBA addressed the problem. Sharing your findings with the press may also result in welcome publicity and awareness building.

survey

The Scope

The first step is deciding the scope of the survey. It can range from a few questions to multiple pages.

Alternatively, you can hold focus groups with representative clients. Some organizations do both, gathering baseline data and direction from a survey and then taking an intensive look at key issues with a focus group.

The two basic types of surveys are census and sample.

Census means surveying the entire population group. If it is feasible to survey all your clients, for example, at intake or exit, then that is a simple way to conduct a census survey.

If the population is too large, or you have no control over whether they take the survey, then you will need to do random sampling. This requires receiving a threshold number to make the results valid.

The site www.surveysystem.com/sscalc.htm has a calculator that will figure out sample size by population.

For example, for a population of 1,000, you need 278 respondents to reach a confidence level of 95 percent, with a confidence interval of 5 percent. For 10,000, you would need 370 for the same confidence level.

The Questions

Be specific about what you want to verify or discover from the survey. Depending on your constituents, needs assessments can cover training needs, community issues, knowledge or skill gaps, preferred service methods and personal goals, as well as barriers to those.

Don’t forget demographics questions – age, gender, income, size of household, education, location – whatever will help you profile the respondents. Then you can compare summary results to your target population to be sure the group is representative.

Sensitive questions like age or income are best asked within a range. And put demographics questions last. That placement will seem less intrusive than hitting them with personal questions right off.

Frame questions in a way that will yield the answers you are seeking. If you don’t want to limit your respondents from sharing their thoughts or opinions, use open-ended questions. Be sure to use simple and clear language. Overly academic or formal wording may confuse participants.

Sometimes yes and no questions are appropriate. They often remove the ability to gather shaded responses. If your question can be answered, “It depends,” use a multiple choice or opinion range, called a Likert scale. Add a “not applicable” choice to prevent false low rankings.

To zero in on priority, ask respondents to rank the top three answers. Allowing them to choose “all that apply” is a way to gather a lot of feedback from one question.

Once you’ve developed your set of questions, test your draft survey. Ask one or more clients to take it and give you feedback about how user-friendly it is. The results will show you if the question format gives you the information you need.

The Format

Also consider your audience when deciding how to administer the survey. Is an online survey or paper survey more likely to be answered?

Many groups use Survey Monkey as a tool for online surveys. You can also use the program to input answers manually from paper surveys, which simplifies collating.

If you have more than 10 questions, you will need to pay for a subscription. Survey Monkey’s analysis features are limited, but they are probably fine for most basic surveys.

Digging into the data for relationships between more than two factors will probably require professional software like SAS or SPSS. Survey Monkey subscribers have the ability to export results into SPSS for further cross-tabulation.

Focus Groups

If you hold a focus group, try to invite up to 10 people who represent a cross-section of your clients. Using a professional facilitator will be perceived as more objective with less likelihood of answers influenced by existing relationship.

Keep the setting informal and relaxed. Make it fun. Recording the session will help capture thoughts and input that might otherwise be missed. Questions should be open-ended and general. From there, the facilitator can dig into the details.

As an alternative, one-on-one interviews can be held to shed light on the survey results. These are time-intensive but can yield much information. One-on-one removes any “group-think” responses.

Once you have the results, the fun of analyzing begins. Hopefully, you’ll find that your organization is on the right track.

You may find ideas for new and better ways to serve clients. In any case, you will have built the relationship because most people enjoy sharing what they think and are glad to be asked.


We are now finishing our third audit with GRF.  In particular, Andreas Alexandrou and Dino Karamalikis have kept us on a track of continuous improvement in our financial processes and office management.

Wendy Gilbert |  Director
MarylandOnline