Introduction
Presenting the Data
Verbal Description of Data
Numerical Description of Data
Conclusion
References

Related Factsheets

Local Program Evaluation
Developing Evaluation Questions
Evaluation Measures

Outcome Evaluation Manual

Evaluating and Reporting Outcomes: a Guide for Respite and Crisis Care Program Managers

Introduction

The evaluation "report" is one of the most important parts of the evaluation. It is the official record of the evaluation. For many people it is the only part of the evaluation that is seen. A good evaluation "report" answers some or all of the evaluation questions. An evaluation "report" does not have to be a written report! However, there should be a written document available to the agency.

There are many ways of communicating results of an evaluation including:

• informal communications (in the hall, over coffee, before and after meetings);
• less formal meetings (small groups, no minutes);
• formal meetings;
• memos;
• written reports;
• newsletters;
• electronic communication;
• visual presentations (slide show, video show);
• public meetings; and
• local media presentations.

A good evaluation report:

• describes the program and the steps in the evaluation;
• explains the procedures used;
• presents the findings;
• draws conclusions about those findings;
• and is prepared with the audience in mind.

More than one report is frequently useful. A report is often used for different purposes, and with different audiences. Many people do not need a long report, or even a written report. A short summary report, or a verbal report, will satisfy the needs of such an audience. On the other hand, some people need the detailed information which can only be provided by a long report. There is also reason to have a more detailed report for archival purposes.

Presenting the Data

In general, most evaluations conducted by local programs would lend themselves to descriptive analysis of data. Descriptive analysis is a way of summarizing and aggregating results from groups. If an evaluation has been conducted which employs a control group, or measures changes in program participants over time, then it might be appropriate to employ inferential analysis in which a decision is made about whether the particular results of the study are "real". More emphasis will be placed on descriptive analysis in this fact sheet.

Verbal Description of Data

Many reports rely on narrative information to present most, if not all, of the necessary information. Narrative information may be presented in three ways: standard writing style; tables; and/or, figures, diagrams, maps, and charts.

Standard writing style, that is, the use of sentences and paragraphs, is often the best way to present information, especially to audiences that are not accustomed to working with charts, graphs, tables, numbers, etc. It is the only way to present information such as examples and explanations. If standard writing style is used to summarize the results of open ended questions ("What do you like most about the program?"), it is often useful to give some indication of how often a particular response was given.

Tables represent narrative or numerical information in tabular fashion. A table arranges information in rows or columns, so that data elements may be referred to easily. They provide a clear and succinct way to present data, and are often more simple and understandable than standard writing style. They also facilitate the interpretation of data.

Figures, diagrams, maps and charts present verbal information visually. They often describe information more clearly than several paragraphs of description. Common forms of figures are: flow charts; organization charts; GANT charts; and/or maps.

• Flow charts are particularly useful for presenting relationships and/or describing the sequence of events and the location and result of decisions.
• Organization charts are useful for presenting the chain of responsibility in a program.
• GANT charts list a set of tasks. They indicate the time each task is to be performed and by whom.
• Maps visually describe certain geographical areas. They are useful in describing different conditions for individual geographical areas.

Numerical Description of Data

Data are not only described in narrative, they are often described numerically. Three of the most basic types of summarization are:

• frequency distribution;
• percent; and
• average.

Each of these types of summarization may be presented as part of the text or arranged in tables or figures (graphs). Inclusion as part of text ("The average age for children served was 18 months") is an obvious way to report data.

Frequency distribution determines the number of units (e.g., people) which fall into each of a series of specified categories. In order to do a frequency distribution one must have categories. Reporting on age, for example, requires that you group the data first before constructing a frequency distribution (e.g., "birth to 2 years," or "3 to 5 years"). The evaluation might look to see how many parents were members of particular racial or ethnic categories, how many were known to protective services, or how many were referred from a range of referral sources.

Frequency distributions are not limited to descriptions of consumer characteristics. Program activity can also be presented categorically. The frequency distributions can be presented as tables or graphs (usually bar graphs). See figures 2 and 3 for examples.

Percent is another useful way of describing data. A frequency count can be converted to percent by dividing the number of units for a particular category by the total number of units and multiplying by 100. Percents are often more easily understood than the corresponding frequency counts. Percents can be represented in the same manner as frequency counts. In addition, a pie chart is useful in breaking the total group of people into the percentage of the total represented by each category. See Figure 4 for an example.

Figure 4: Crisis Nursery Percents

An average is a way of summarizing all of the information into one number. It can be used with data which is non-categorical numerical data. You cannot have a numerical average for gender or race, for example. Using a numerical average is very powerful, but it can also be misleading. A few data points which are very different from the others could substantially change the numerical average. For example, if the ages of children you serve are generally between 1 and 3 years, but you get one child who is 18, the average may be thrown off. Averages can be represented in tables or graphs.

Conclusion

There are some things which can be done to encourage the utilization of evaluation results:

• Write a good report which attends to the important issues which were raised in planning for the evaluation.
• Write a report which pays attention to the audience's needs and abilities.
• Make concrete and usable recommendations.
• Involve the program staff in all steps of the evaluation process.
• Disseminate the results in various ways and in various settings.
• Write a report which can be easily understood by those who receive it.

References

Fitz-Gibbons, Carol Tayler (1987). How to Analyse Data. Newbury Park, CA; Sage Publication.

Morris, Lynn Lyons (1987). How to Communicate Evaluation Findings. Newbury Park, CA; Sage Publication.

About the Authors: David B. Langmeyer, Ph.D., is an Evaluation Consultant to ARCH. He was Chief of Evaluation and Research in the NC Division of MH/DD/SA Services for thirteen years.

Gail S. Huntington, Ph.D., is Director of Evaluation and Research for ARCH. She is a Research Investigator at the Frank Porter Graham Child Development Center at the University of North Carolina at Chapel Hill.

ARCH Factsheet Number 15, August, 1992