Using Data for School Improvement:
A Guide for Principals

Schools and their principals are buried under data. Test results arrive in formidable packets or, often more intimidating, on CD-ROM or computer disks. Attendance summaries, budget sheets, medical information, permanent records data, police reports and dozens of other kinds of data appear almost daily, presenting the principal with the daunting task of even storing all of the information, let alone digesting it and using it for school improvement.

At the same time, our professional intuition and most research on school improvement tells us that these data can provide valuable information to help in the complex task of school improvement. And therein lies the key question: how can principals turn this mountain of data into useful information?

Fortunately, the process of using data is usually not as complex as storing, retrieving and analyzing it – tasks that are often performed for us by testing companies, state departments of education, or district management information departments. In fact, using data effectively generally starts with answering several very important questions:

1. What data are available to us?
2. How can we turn these data into information?
3. How do we use this information for school improvement?

Data or Information?

To begin, it is important to distinguish between data and information. Data is simply the product of some collection scheme. It might be test results, raw attendance figures, or a budget sheet. By itself, it tells you very little. Indeed, if a principal tells us that his "average daily attendance rate" is 88%, we may believe that is pretty good, unless it was 92% last year, or the district average for similar schools is 95%! Digging a bit further, we may find out that 88% is the attendance rate at noon. If attendance were taken at the beginning of the school day, it might be much lower – or higher. By itself, that bit of data tells us very little about the school and its performance.

Test scores suffer from similar problems. If 58% of the students reach proficiency on a particular state test, we still know very little about the school’s achievement unless we know its own achievement history, how it compares with other schools in the district, or what the state norms are for schools of a similar type.

Such test score results may also mask very specific problems. If 58% of the students achieve mastery on an Algebra test, for example, the very poor performance of some students (girls, minorities, and limited English proficient kids) might be masked by the very high performance of other groups. So, while the average performance of the school might be 58%, that may result from 100% of one group mastering the test while only 40% of another group achieves such mastery.

Other data are even more slippery. Drop out rates can be calculated in a number of ways. Among them are the students who begin 8th grade but never finish high school, the students who officially "withdraw" from school, or the number of students who miss more than some specified number of consecutive days without a legitimate excuse. The first set of data may yield a very high figure but not account for students who move from the school’s attendance zone. The second set is likely to under-report the problem since relatively few drop outs show up to let you know they are not coming back. The third set of data may present more accurate information, but only if we know where those students really are. All of this is complicated even further by our inability to detect how many students stop attending high school but ultimately earn a GED. So, even under the best of conditions, conventional ways of measuring "drop outs" may be misleading, and may, indeed, tell us only who has interrupted their formal, school-based education for some period of time.

All of these issues suggest that using data for school improvement requires attention, thought, and some planning. After making an initial investment in setting up a data and information management process, however, the amount of effort devoted to maintenance can be minimized and still produce useful information.

What is Available?

A first step toward using data that is useful for school improvement is to identify what is routinely available and how it is delivered to the school. As a practical matter, the most important consideration here is to identify all of the data that are related to the school’s major goals and give less attention to information that is peripheral to those goals.

For example, if the school’s major objective is to improve achievement on state tests, that test information should be the focus of data collection as well. If a secondary objective is to improve attendance and on-time arrival at school, those data become significant too. It is important to focus data management efforts on the most important aspects of the school’s agenda. Otherwise, it is much too easy to become overwhelmed with all of the data that are available.

Once the major information sets have been identified, it helps to make an inventory of what is available for study and analysis. For state testing programs, the results probably arrive as a report on individual school performance compared with other schools and districts. It may also contain individual student results that range from simple "pass-fail" designations to detailed descriptions of individual student performance, including individual items or objectives that they have mastered. Obviously, the most useful data for school improvement is that which gives the most detail on individual student performance.

Consider other data sets as well. Do teachers give practice tests? Do you have access to student performance data from prior years? Do you have information on other testing programs in the district? Are specific achievement test results (e.g., reading, other standardized tests) available as well as the state tests? If so, all of this may help to give a bigger achievement picture than can be determined from a single test.

The same process is useful for attendance data as well. Obviously, the school keeps attendance data as required by law and policy. That provides a good overall picture of the attendance issues in the school. Other data can be equally useful, though. Teacher attendance records may show patterns of absence or tardiness at particular times of the day, in specific subjects, at different times of the year or in various parts of the building. Generally, more specific information allows school leaders to target interventions in ways that have the most promise for success.

Once your inventory is complete, you may find that there is other information that would be useful to the school’s improvement efforts. If so, this early phase is a good time to involve teachers and other school staff in developing new data sets to fill in the blanks in your information profile.

How to We Turn Data into Information?

Turning data into information simply requires that it be put into a form that makes sense to people in the school. Knowing that the school has a 58% mastery rate on the state's algebra test is not useful unless we also know that it is higher or lower than in the past, that it is equally distributed over the different groups of students in the school, or that it is comparable with other schools with similar composition. To turn data into information, several approaches are useful.

Disaggregate. Taking data apart helps to reveal strengths and weaknesses that are not evident when it is aggregated together. In the example noted earlier, an overall score, such as 58% on the algebra test, may mask the very high or low scores of specific groups or individuals. Disaggregating data by gender, ethnicity, language proficiency, special program status, or a host of other variables can help principals identify targets for school improvement efforts.

Secure Expert Commentary. Many times, experts in a field can tell principals if their data deviates in some meaningful way from normal trends or if specific data patterns call for specific kinds of interventions. As an example, females typically outperform males on most academic measures in elementary school, but the trend is often reversed in high school. This pattern may point to a specific problem in the curriculum or in teaching methods. A deviation from this pattern may indicate that the school is doing something very right that it will not want to change. Often, a few hours spent with someone who knows a field well will provide tremendous insights about what the data are really saying.

Don’t ignore the experts in the school, either. Math teachers who are presented with the data on their students' performance may be able to offer very insightful commentary on how the curriculum is structured to produce exactly the results the school is getting -- whether it is what the school wants or not. In fact, a routine practice in using data for school improvement should be to ask teachers to comment on the results and provide reasons for the outcomes the school is attaining.

Analyze Against Curriculum Goals. Most state tests and many standardized tests are directly linked to curriculum standards in each field. A very useful process is for teachers to look at the curriculum standards from their field and, using test data, determine how well a standard is being met by the students. The tools for such an analysis are quite simple; they consist only of a list of the state curriculum standards and a set of student test results by item and test objective. By matching the students' performance on specific tested objectives with the curriculum standards, teachers can identify parts of the curriculum that need more or less attention to improve test performance.

How Do We Use Data for School Improvement?

Once the raw data have been put into a form that can be understood in the context of the school's program, the effective principal must get the faculty to focus their attention on this information and incorporate it into their planning. Among the most useful strategies for bringing these data into the planning process are these:

Discuss It. Reserve a specific time during faculty, department or team meetings to talk about data and examine its implications for teaching and learning. "Chunk" the data so that it is not overwhelming and teachers can think about it as they plan curriculum and teaching.

Take Credit and Responsibility. One principal uses a very powerful strategy to cultivate a sense of responsibility for the data yielded by student testing. In one faculty meeting, he presents a list of the strengths in the data -- a list of what the students did well. Then, he asks the faculty to explain those results. Naturally, most of the strengths can be attributed to the school's program, the fine teaching in the building, and special emphasis given to specific topics in the curriculum. The next week, he presents a list of the weaknesses and conducts the same discussion: "Why are we getting these results?" He has found that if the staff accepts the credit for the students' strengths, they are much more likely to accept responsibility for the weaknesses, so genuine instructional planning can be pursued.

Set specific targets. Rather than "raising test scores," effective leaders focus on specific goals that can be relatively easily attained. So, instead of "improving reading," a school may set the goal of "improving students' use of context clues for vocabulary development." Then, everyone in the school can focus on the same skill or concept in the improvement effort. In one particular case, the reading specialists provided lesson plans that could be adapted to every subject and asked every teacher to teach the skill at least three times in one month. The result was astonishing: students received instruction as much as fifteen times in one month -- three times in each of five classes! It is impossible to leverage that much instructional time for basic skills unless everyone is working together on specific targets.

Focus on the things you can change. Too often, educators become discouraged by the overwhelming social and community conditions that affect their students. They may believe that their influence is small compared to the challenges of everyday life that their students face.

One Florida principal uses what she calls "the Geometry of Achievement" to focus her faculty on the things they can change. She explains that the hypotenuse of a triangle represents student achievement. One leg of the triangle represents all of those things that are beyond the control of the school -- economics, family circumstances, community problems, and so on. The other leg represents the things the school can change: curriculum, instruction, student support, and a host of other variables. She then asks the faculty how to make the hypotenuse of a triangle longer. It's not long before one of the math teachers says, "by making the legs longer." "Suppose you can't alter the length of one leg," she asks. "Then you have to make the other leg as long as possible," comes the logical reply. Obviously, the moral of the lesson is that if you can't change everything, at least focus on the things you can change to promote student growth.

First Steps

Using data for school improvement begins with a decision to rummage through the daunting array of data that enters the school and selecting that which is directly related to school improvement goals. Once the process has begun, tenacious principals have found that their staff begins to take on more of the responsibility for turning data into information, mostly because it makes it so much easier for them to focus their precious instructional time and effort on the things that really matter.

Resources

For more help with using data for school improvement, visit these useful websites.

School Reform by the Numbers. This Kentucky schools uses data in routine planning at all levels: school, team and individual teacher. Their model transfers well to other schools that have state mandated testing (PDF).

Creating a Standards Based Classroom. Also from Louisville, this interesting article shows how to plan effective and engaging instruction based on state standards (PDF).

Research Based Strategies to Achieve High Standards: A Guidebook for School-Wide Improvement. "Getting Results: Gathering Information for Informed Decisions." From West Ed, developer of the U. S. Department of Education's Comprehensive School Reform Demonstration Project, this excellent chapter outlines how to gather and use needs assessment data for use in school reform and instructional improvement.

Data Inquiry and Analysis for Educational Reform. From ERIC, this succinct document outlines what kinds of data to collect and how to use it for meaningful school improvement.

Resources for Principals and Curriculum Leaders from Middleweb. This excellent collection of resources provides helpful guidance on all aspects of curriculum reform, including the use of data to drive reform initiatives and improve instruction.

to the top

Prepared by Howard Johnston