Section 9. Developing Baseline Measures

 

 

So you've set out to identify the problems that exist in your community and you've started collecting information about a particular problem. How will you get started attacking the problem unless you first have some idea of its extent and intensity? Once the intervention is in place, how will you know how much things have changed unless you know the extent of the problem before it started? This is where baseline measures come in to play.

What is a baseline?

To make a change in your community, one of the first things you'll need to do is figure out how much the different factors and trends you're examining are happening in the first place. Try to find out how prevalent any problems and positive tendencies are, how often things happen, the duration and intensity of most incidents, etc. The things you keep track of in order to obtain this sort of information are called baseline measures. In other words, the baseline provides a starting point against which you can compare subsequent changes during and after your program. Baseline measures can also be shown in graph form to make changes over time easier to see.

Here's how baseline measures work. Suppose you've observed a high rate of pregnancies among adolescents in your community, and your organization decides to do something about it. So you gather statistics over a one-year period about the number of reported pregnancies among adolescents, and then you measure again, comparing the new figure against your baseline. Depending on whether the numbers have gone up, gone down, or stayed about the same, you'll know whether conditions have changed since your intervention began. The comparison alone can't prove that your intervention caused the change, since other factors may also have contributed, but it provides important information for evaluating your efforts.

Let's look at some other examples of baseline measures from everyday life.

Examples: Baseline measures in everyday life

• A parent hears their baby crying. They know from past experience how and when their baby usually cries -- that's their baseline. Because of that, they can tell this time around whether the baby is crying just because they're fussy, or whether there might be something to worry about.
• Rafael gets up in the morning and doesn't feel well. He makes that judgment based on knowing how he usually feels when he gets up. That usual feeling, based on past experiences, is his baseline. He uses it to help decide whether he should get up and go to work, or call in sick and get some rest.
• Your friend Marisol tells you you're looking good. She knows how you usually look-- that's a baseline for her. If you're looking particularly rested, or fashionable, or bright-eyed this morning, Marisol notices and comments.

All of these simple everyday examples involve baselines. Sometimes people may call them other names: reference points, adaptation levels, anchors, or norms, for example. But whatever we call them, we all have them and we all use them. In a way, they are essential for all of us in making judgments about people and things. Without baselines, it would be much harder to navigate our way through the world.

Most everyday baselines are casual and informal. We may not even notice them; we certainly don't measure them. In making decisions about community actions, and about public policies in general, however, we sometimes take baselines a lot more seriously. For example, every month, the Consumer Price Index is released by the Bureau of Labor Statistics. The numerical value of that index, which helps track changes in consumer prices over time, relies on a reference period. Likewise, many other indices, including stock indices (e.g., the Dow Jones), local quality of life indicators, et cetera, rely on reference points and measures of change over time.

Comparisons with baseline data are an important source of evidence for policy and program decisions. If a new law has raised the penalties for alcohol-impaired driving, has alcohol-impaired driving decreased? What kinds of activities will draw visitors to local parks, based on data we already have on park use? Good decisions in these cases and countless examples more depend on good collection and use of baseline and follow-up data, along with consideration of other factors that may influence the results.

Establishing baseline measures is often an early step in program planning and evaluation, followed by implementation of the program and continued measurement and analysis of what happens over time.

Why use baseline measures?

Baseline measures can help you assess whether your efforts are working.

To plan a truly effective program, you have to know how much of an effect your efforts may be having. You need to have an idea of the level of the problem before your efforts begin in order to assess what changes afterward. Recording baseline measures, which you can then compare with whatever the numbers are after your intervention has started, will help you figure that out. Other influences should also be considered before concluding that your intervention caused the change.

A baseline can help you make sense about something that might be too massive and complicated to understand otherwise.

A question like How well are our schools working? might be overwhelming to try to answer. However, keeping track of baselines, such as standardized test scores or high school graduation rates, can help you better understand part of the bigger picture. No single measure will usually tell the whole story.

A baseline can help you decide whether this is a good time to start an intervention or whether a particular intervention is appropriate.

Say you're working to decrease fatal car accidents in your county. One of the ways you're thinking about doing this is to start a program to encourage seat belt use. Getting some idea of how many people in your county are consistently using their seat belts will help you decide whether you should spend any time and resources on such a project. The rate of seat belt use will be your baseline measure. If 98% of local residents are already using their seat belts most of the time, you may want to explore other possible interventions.

Baseline measures can sometimes tell you if an intervention may not be necessary or if the problem has been misunderstood.

For example, community leaders may be concerned about an increase in gang-related activities among youth and demanding programs to discourage it, but good, accurate baseline measures of reported incidents or youth involvement could show that the perceived increase isn't supported by the available data.

Baseline measures can help you identify methods that may not be producing the intended results.

If there is little or no change compared with the baseline, that's a reason to look more closely at what's happening. The intervention may need more time, may not have been implemented as planned, the measure may not adequately capture the desired change, or the method itself may need to be adjusted.

For example, let's say you're working to increase the numbers of people who are pregnant in your city getting prenatal check-ups, and you've decided to use a series of public service announcements to do this. By comparing the number of people receiving prenatal check-ups after a given period to your baseline measure -- the number of people receiving prenatal check-ups just before the public service announcements started running -- you can see whether the numbers are moving in the desired direction. Maybe the numbers are increasing and you'll decide to continue running the public service announcements, or they may be remaining steady or even decreasing, in which case you might want to look more closely at why and consider adjusting your approach.

Keep in mind, however, that your method or intervention may take some time to produce the desired effect. Behavior and community-level change may not show up immediately. Be sure to allow enough time before concluding that a method or an intervention isn't working. It could be that it just needs more time, or that other factors are influencing the outcome.

How do you develop a baseline?

Pick an indicator or indicators that best reflect the behaviors or conditions that are most important to you.

An indicator is something measurable that can be used to identify a change in trends or conditions. An indicator can be the number of alcohol-related car accidents per county per month throughout your state, the number of people requesting a particular pamphlet that your organization distributes, or the number of pregnancies among adolescents in your community in a year.

The indicator needs to be relevant: it should tell you what you need to know.

Ask yourself these questions:

• Does this represent what's most important and pertinent to our community?
• Does this show some facet of the long-term well-being of our community?
• Is this measure showing what it's supposed to measure and not some by-product?
• Can this measure be compared to progress in similar communities on this issue?

Below is an example of a group selecting a baseline measure to use in tracking the effectiveness in one of its programs.

Example: A gang-prevention and youth engagement project for urban high school students

Some indicators that you might want to consider using as your baseline measures could be rates of:

• Students involved in extracurricular activities and faith-based groups
• Substance use
• Students reporting gang involvement or recruitment pressure
• Students in counseling programs
• Students in conflict resolution programs
• After-school and weekend employment

If one of the things your group is planning is an after-school discussion group, then the first indicator would probably be a good one for you to use as a baseline measure, because it should give you a good idea of the numbers of students who might be likely to participate in an after-school activity.

Find measurements on those indicators.

Once you've chosen indicators, decide exactly what you're going to measure, and for how long.

For example, will you measure violent gang-related incidents on school property during the school year? Will you measure the number of alcohol-related automobile fatalities over a four-week period? It's possible that someone else has already measured these things; if so, then you'll just need to verify (and, if necessary, update) the information. Otherwise, you or someone else will need to go out there and measure them.

Some things to consider if you're doing the measures yourself:

• What characteristics of behavior or conditions should be measured? Some of those characteristics include frequency, rate, percentage of occasions something occurs, and duration.
• Under what conditions should you collect data?
• Will you observe continuously or do sampling? If sampling, how often to do so?
• Will you collect data for given periods of time, or by intervals? If for given periods of time, how long will those periods be? If by intervals, how long will the intervals be?

Remember that a useful baseline often includes information gathered at several points over a period of time rather than simply a snapshot of information gathered over, say, a single weekend. The amount of baseline information you need will depend on the measure, how much it normally varies, and what you hope to learn from it.

How do you interpret changes from the baseline?

Let's say that you now have data for your baseline measure, as well as data collected at a handful of different times afterwards. How do you make sense of this information? First, you should know a bit about the different types of baseline data patterns.

When you present your baseline measures in graph form, you can learn a lot about the pattern and extent of a problem in your community and get useful information for deciding how to plan and evaluate an intervention.

Example: Baseline data patterns

Stable baselines

With a stable baseline, there's no evidence of upward or downward trends; things may fluctuate a little over time but for the most part the data points fall into a pretty tight range.

A stable baseline can make changes after an intervention begins easier to interpret because there was no strong upward or downward trend beforehand. Even with a stable baseline, however, other events or influences may contribute to later changes, so the baseline by itself does not establish that the intervention caused them.

Ascending and descending baselines

As you might guess from the names, an ascending baseline means that whatever is being measured has steadily increased over time, and a descending baseline means it has decreased.

Unstable or variable baselines

When the data points range all over the place and there are no clear trends, you have an unstable or variable baseline. A variable baseline can make it harder to tell whether later changes are related to the intervention. Rather than automatically delaying action, you may need to gather additional baseline data, examine why the measure varies, use additional indicators, or choose an evaluation design that takes the existing variation into account.

When you have a more complex baseline, you need to give some serious consideration to whether the health problem is increasing, decreasing, or fluctuating and what those patterns mean for your planned intervention.

Say, for example, that the baseline measure you're graphing stands for the rate of smoking among adolescents in your county. If the baseline is descending, indicating that fewer and fewer adolescents are smoking, that trend is important to consider when deciding what intervention is needed and when interpreting later changes. If the baseline is ascending, showing that smoking among adolescents is on the rise, that may add urgency to the need for action. In either case, the pre-existing trend should be taken into account when evaluating what happens after the intervention begins.

How do you use baseline data to develop an intervention?

Decide which problem(s) to address

Based on the data, decide what problem(s) should most be addressed by your group or coalition. What looks like it most needs to be dealt with? Is it something you can reasonably expect to be able to change?

Identify priority groups for the intervention

This means deciding who or what your intervention will focus on. A given group of people? An institution? A policy or environmental condition? Consider how behaviors and surrounding conditions contribute to or maintain the problem. Figure out what your research results suggest about relationships between the problem(s), the priority group, and the conditions that shape the issue.

Develop an action plan

• Set a behavioral goal for the intervention
• Study what knowledge and information you already have about the community and the problem to decide what procedures to use.
• Decide who will be responsible for leading and supporting the planned changes.
• Determine how to involve those affected by the problem(s) in the solution.
• Study other models of change. What existing methods for dealing with the behaviors or conditions have been successful in similar communities?
• Decide how you will go about explaining the intervention to the public.

Get going! With the knowledge you've gained from checking your baseline measures, you will have a much better chance of making informed changes in your community, and you can use those measures to monitor your success.

In Summary

Using baseline measures to figure out how prevalent problems and positive tendencies are in your community can be very effective in helping you monitor change and assess how your efforts are progressing. By giving you a starting point for comparison, baseline measures can be enormously helpful to planning, monitoring, and evaluating your programs.