Skip Navigation
The Forum Guide to Data Ethics
NCES 2010-801
March 2010

2. Appreciate that, while data may represent attributes of real people, they do not describe the whole person

Advising kids about their future educational opportunities and career choices was Mrs. Johnson's passion. Some of the students in her middle school had the ability to do just about anything they set their minds to, and she was always sure to tell these kids that the sky was the limit. But with other students, well, that was another story.

Colette, an eighth grader with a history of barely passing math courses, walked excitedly into Mrs. Johnson's office and told her that over the weekend she had taken an elevator ride to the top of the tallest building in the city. Clearly the experience had changed Colette's vision of her future. "I've decided to become an engineer who builds skyscrapers!" Mrs. Johnson quickly reviewed Colette's math scores, attendance history, and family situation, and decided that the data suggested another path. "You know, Colette, becoming an engineer requires a lot of specialized education, and college is very expensive. Maybe you shouldn't set your sights so high. I like your haircut. Have you ever thought about cosmetology?" Collette looked dejected. "Are you telling me that I can't build skyscrapers? What if I started studying really hard?" Mrs. Johnson decided not to sugarcoat her analysis. "No, dear, higher math can be quite challenging and your academic record isn't very strong." Colette chose to stand up for her new dream, "But I haven't even gotten to high school yet, Mrs. Johnson. Couldn't I take geometry next year and improve my grades?" Mrs. Johnson smiled condescendingly and said, "That's very ambitious, Colette, but geometry is much more difficult than it sounds. Why not register for general math and use the extra time to get a part-time job so that you have more experience when you start looking for a job?"

There are limits to how well data can portray people—who have complex thoughts, needs, and emotions. Data collected in schools are indicators of uniquely personal events, conditions, outcomes, and ambitions. Each piece of data in a student or staff database represents a particular attribute of a real person, and may be used to make important, lasting decisions. The data may be dry facts to the people who look at them, but to the people whose educational and professional experiences are being recorded, they are the pieces of a very personal academic career and individual history, recording their achievements and disappointments.

However, teachers, counselors, administrators, and others who use student records should be careful to focus on the student sitting before them as well as the student's history documented in a database. Some information could be incorrect, if data integrity is not maintained. Some could be misleading, if taken out of context. Did a disciplinary incident occur when the student was facing problems at home? Does the new student who's doing so poorly in math need to have his vision checked to be sure he can see the chalkboard?

And, perhaps most important, people can change. After all, that's the point of education! The double warning in this canon is that data cannot represent the total person whose life they record, and they cannot predict with absolute precision what that person will become in the future.

Each piece of data in a student or staff database represents an attribute of a real person, but these data cannot adequately portray all aspects of a multifaceted individual.

Recommended Practices and Training

  1. Accept that there are limits to how well data can describe people—people with complex thoughts, needs, and emotions; people with physical or psychological challenges that may not be well understood; or people who, through no fault of their own, live in circumstances that are unhealthy, unsafe, or unstable.
  2. Be especially careful about making personal or professional judgments about people based solely on data. Be particularly alert to data that may be flawed, narrow in scope, or otherwise of limited applicability.
    1. Just because data can be used to answer a question or inform an opinion does not mean that the information is entirely accurate, reliable, and unbiased.
    2. Be very cautious about using data for purposes other than their original intent. Be sure that doing so does not violate individuals' right to privacy or any agreements of anonymity that you or your agency has made. Aggregations of data may be published if personally identifiable information has not been disclosed.
    3. Effective, data-driven decisionmaking draws from multiple sets of data that support the same interpretation. Do not draw from a single source, if at all possible, and look at data from multiple sources over time to see if the findings are consistent.
  3. Be willing to challenge commonly held assumptions and prejudices related to descriptive data.
    1. For example, do not equate disability status with decreased intellectual aptitude or potential. In some cases, disability status reflects variation in learning styles rather than academic capacity, and some students with disabilities do not show differences in their ability to function in a school or life setting. In other instances, accommodations may permit students with disabilities to function at high academic levels.
    2. Do not automatically equate school success with life success. Academic success is important, especially within the context of the education system, but people can find happiness, prosperity, and success in life without being the highest achiever in school.
  4. Train data handlers to understand the limitations of data as a tool for describing individuals and categories of people by age, gender, racial/ethnic group, language of origin, job type, and other categories. Customize training efforts by job type as appropriate for communicating concepts and translating instruction into practice. Lead a discussion in which you ask if anyone has ever been treated unfairly because of data in a school or work record. Alternately, lead a discussion in which participants are asked to describe a situation in which outcomes showed the data about them to be poor predictors of success. Another activity is to provide participants with a scenario and ask them to fill in the missing parts. For example, prepare a short hypothetical resume that shows a break of several years in work history, or a string of jobs lasting no more than a year, and ask "applicants" to explain these to a job interviewer. See if the "off the record" information changes the interviewer's opinion of the applicant's credentials.