Data Sources and Certification

Institutional Data

Institutional data is a very broad term for any piece of information that the university obtains as part of its normal operating procedures. These data constitute personally identifiable information and can be tied back to a particular student, faculty, or staff. According to the Family Educational Rights and Privacy Act (FERPA), personally identifiable information can include the student’s name, address, social security number, birthdate or any other information that can be used to directly identify the student.  Given the scale and broad reach of university operations, it is very easy to see the wealth of knowledge that can be gained from institutional data. However, one must exercise caution and carefully consider the appropriate uses of such data in research – even if they are readily available when performing the daily duties of one’s job.

Types of Data

When asked what sorts of information the university collects about students, the most common answer involves some sort of enrollment data – course registration history, grades, or majors. The scope of information that the university collects on any given day, however, goes well beyond this. Data are collected on student financial records, internet access, card swipes, and participation in extracurricular activities. Information is also collected more directly in the form of survey responses from students. All of this (and more) constitute institutional data.

Risks in Data Analysis

When doing practical ethical data analysis, ensure that you are accessing any personal student data through a secure network, whether on a secure campus machine or a VPN if not on the Purdue Airlink system.

Never download data directly to a machine that is not physically secure, even if using your password-protected laptop. Consider accessing the data directly through a Purdue-supported file storage system (as of 2022, Purdue uses Box for its secure data storage software).

Never use a USB or other portable hardware to store data, even temporarily. Each employee is personally responsible for their stewardship of Purdue data.

Validity and Use

Purdue collects a staggering amount of personal educational data. When exploring larger datasets as part of research projects, there are several considerations that will promote ethical practice.

The Belmont Report emphasizes three basic principles:

• Respect for Persons: Research should respect participants’ dignity and autonomy, with particular protections in place for those who have diminished autonomy (e.g., children).
• Beneficence: Research must minimize the risks of harm to participants and maximize the benefits.
• Justice: Research should distribute the benefits and costs of research fairly.
  

Research does not involve sifting through the data, but posing a specific question that can only be answered by analyzing a particular set of data. The data analysis should only answer that particular question, not simply as a “pool” of information where correlations can be run until finding something of statistical significance.

Ethics in Institutional Data Use

Transparency

Students are significantly unaware or under-aware of the use of their personal data for institutional data analysis purposes.

Justice and Fairness

Students are generally accepting of the use of certain types of institutional data for helping them succeed academically, support their learning, or avoid low grades, but they are not in favor of using data for comparative purposes.

Responsibility

Acting with integrity is generally the most common application of responsibility in data use and analysis. However, most people tend to connect integrity with individual action and “not getting one’s hands dirty.” When conducting ethical data analysis, responsibility also implies a duty to question requests from superiors for analysis that might violate those ethics. In short, responsibility does not end with our own commitment to conduct focused and professional searches, but to ensure our organization does not waver in its commitment to student justice and transparency when analyzing identifiable data.

Privacy

In contrast to the European Union, the United States and the state of Indiana offer minimal to no formal guarantee or protection of private data. While the limit of the law does not restrict institutional data usage, we are still bound by FERPA as stewards of the data and any privacy breaches. Privacy has different cultural connotations, especially when comparing US/Europe to East and Southeast Asia. however, the general understanding of privacy usually involve an individual’s ability to define and limit access to personal information. One of the easiest ways we can integrate privacy into practice is by limiting our involvement with identifiable data. If possible, ask your unit’s data steward or the campus data steward if you can access data without personal identifiers, while retaining a coded, unique identifier for each member of the data set.

Specific Data Sources for SoTL Projects

There are many potential data sources that you might consider using for your SoTL projects. Keep your research question in mind when deciding what data you will be collecting for your SoTL project: while it can be tempting to collect as much data as possible, focus on what data will help you answer your research question.

Registrar Data

The Registrar collects and organizes an abundance of data on students and courses. These data include student demographics such as gender, ethnicity, age, transfer student status, international student status, underrepresented minority student status, first generation student status, and major. These data also include academic variables such as course grade, overall GPA, term GPA, credit hours and standardized test scores (e.g. SAT/ACT, TOEFL). Financial data, such as family income, is not available through the Registrar and is difficult to obtain for SoTL projects.

Course Data

Much data are collected throughout each course for normal educational purposes. These data include assignments, quizzes, exams, class activities, group work and participation. In addition to looking at total scores for many of these, you can use rubrics or exam breakdowns to look at specific questions or learning outcomes that might be related to your research question. BrightSpace can also provide data that might be relevant including whether students have accessed certain course materials or watched course videos. If you include mastery quizzes, you could include the number of attempts that students make for each quiz. Finally, your own notes or observations can be used as data for your project if relevant to your research question.

Course Evaluations

Course evaluations can provide both quantitative and qualitative data for your SoTL project. While course evaluation questions tend to be standardized, you can talk to your department about the possibility of adding a few additional questions which might be more targeted towards your research question. The open-ended questions might also elicit relevant qualitative data which could be valuable for your project.

Surveys

Conducting additional surveys with your students can be a relatively easy way to collect some specific data for your SoTL project. Your literature review is a good place to start to identify measures that you might want to use in your surveys. When surveying your students, try to keep your surveys fairly short with a limited number of open-ended questions to improve your response rate. You can also consider providing a small amount of extra credit for completing surveys as long as you have an alternative way for students to earn extra credit if they do not wish to participate in your research project.

Interviews/Focus Groups

Conducting student interviews or focus groups can be another valuable source of data for your SoTL project. Interviews and focus groups tend to provide more qualitative data which might be useful for more exploratory studies. You may want to consider providing incentives for students to participate in interviews or focus groups such as extra credit or food. Also, you will probably need someone who is not teaching the course to conduct the interviews/focus groups so consider collaborating with grad students or colleagues. Alternatively, you could wait to conduct the interviews/focus groups until the course is over.

Here is a Data Collection Worksheet to help you think through this process.