SOTL Research and Human Subject Protection

The Scholarship of Teaching and Learning (SOTL) is a growing area of research. Conducting research on one’s own teaching methods or student behaviors in a classroom is a rewarding line of research, but it can be fraught with ethical and legal pitfalls. SOTL research usually involves intervention or interaction with students and thus falls under the review of the Committee on the Use of Human Subjects in Research (CUHSR or Bradley’s IRB). In addition to human subject interaction or intervention, reviewable research typically deploys a methodology that is systematic and generates results that are generalizable. This means that there is a purpose and hypothesis related to the data collection (systematic) and the data will be analyzed in such a way as to make some inferences beyond the immediate subjects of the research (generalizable). This usually implies an intent to publish or disseminate the outcomes. This type of research must be approved by CUHSR.

FERPA and Research

An important area to consider in SOTL research is the Family Educational Rights and Privacy Act (FERPA), which is a federal law that protects the privacy of student education records. The provisions of the act must be followed when researchers are using student data in their research. The main concern regarding FERPA when doing research is accessing student education records and data without permission or consent. Generally, student education records are confidential and may not be released without written consent from the student. Student educational records are any materials that contain information related to a student and are maintained by the institution. Examples of educational records include graded materials (exams, papers), class rosters, transcripts, and computer screen data. There are exceptions to obtaining consent under FERPA. Please see the Office of the Registrar website for more information and details. (The details of the exceptions can be found in the Federal Register 34 CFR § 99.31.)

SOTL research tends to fall into several broad categories regarding the type of data collected. First is institutional student data, such as ACT/SAT scores or course grades. Another type of student data is generated through normal classroom activities, such as test and assignment grades. These data sets would squarely fall under the FERPA regulations. Finally, there is research data that is collected through the activity of the researcher apart from the course activity, such as a survey. Each of these areas will be discussed.

If your research involves intuitional data, which is data that is officially housed by the Office of the Registrar and is accessible to you, then, if you have demonstrated a legitimate educational interest, you may access that information according to the CUHSR-approved protocol. CUHSR will be looking for 1. a clear legitimate educational interest (defined above), 2. that you will de-identify the data as soon as you can in the process, and 3. that only those on the research team that are Bradley employees who have an educational interest can see the identified data. You will also need to specify in the application the specific data, the time frame of the data, and the specific purpose of the data related to your research and to whom the data would be released to. You will need to carefully describe how you will keep the data confidential and private. Should you need data that is not already accessible to you, you will need to make that request with the appropriate departments. Typically this request can be initiated from the Office of Institutional Effectiveness through a request form. Typically, this data would be released in a de-identified form.

The following content is needed in the consent form about student information:

  1. The specific records or information collected. The requested information should be specified by item (for example, in course ENG 101 assignment on..., final course grade, and exams 1and 2). You cannot request the entire academic record of the course. It must be specific to the research question.
  2. The specific time frame in the data is collected (fall semester 2020 in which you are currently enrolled).
  3. The specific purpose of the data should be clear. You should not collect information that does not directly relate to your research project.
  4. To whom the data is released. This should be to just the professor or the investigator who will de-identify the data.
  5. The consent must include the signature of the student or if under 18, a signature of a parent or legally authorized representative. This should be a written signature or a verifiable electronic signature. Should a student not provide consent to use their data, the researcher is not allowed to use it in their research.

Other FERPA Concerns in Research

Faculty are not allowed to give out a class roster so that a researcher can recruit subjects from that list. This would be a FERPA violation.

Researchers not affiliated with Bradley may not access identifiable student data without permission even though their research may seem to fall under a “legitimate educational interest.” The interest is only legitimate for those within the institution.

Students who are researchers typically do not have a “legitimate educational interest” and thus cannot have access to student information without a FERPA release.

Other Educational (SOTL) Research Not Under FERPA

During normal operations, you can “evaluate” the students’ activities, scores, and metrics without their consent where you have a “legitimate educational interest” within your direct responsibilities at the institution (such as within your course or unit). This is not considered “research” by federal regulations. Nor is it a violation of FERPA if the data is kept internally with those that share the educational interest or if the data is de-identified. However, if you are doing “human subjects research” with the intent to analyze and disseminate the data, then you must have the student's written consent. This is true even if you de-identify the data after you receive it. If you need data from students that are typically institutional data you might consider asking the student to self-report that data, such as a self-reported ACT/SAT score, current GPA, and demographic information. In this case, in your dissemination, you would need to be clear that this data was self-reported.

SOTL Research Not Involving Student Data

Some SOTL research will ask students to do something that is outside the normal classroom activity. A good example is a pre and post survey to determine the perceptions of the student on the effectiveness of a certain educational intervention. This survey would have nothing to do with education occurring in the classroom. Since it is not a part of the normal class, the data gathered is not “student data” and FERPA does not apply. When doing this activity certain elements must be clear (and these apply in principle when also using student course data – see sample consent forms):

  1. The activity is voluntary and the student’s participation or non-participation will not impact the student’s grade or standing in the class.
  2. If an incentive is given, such as extra credit points, then the researcher must give a non-research opportunity with a similar incentive, such as a brief assignment. Thus, the individual who voluntarily does not wish to participate in research has the same opportunity to gain extra credit points. An element of coercion would exist without this non-research alternative.
  3. The students should be explicitly told that the professor/researcher will be blinded to who will participate until the final grade is posted. Thus, students will not have the perception that their grades could be unduly influenced by their participation. Typically, SOTL researchers, if collecting student data and thus collecting signed written consent forms, will have a confidant who is not associated with the research but who is knowledgeable about the research pass out and collect the consent forms, and then hold the forms until the final grade is posted. It would be difficult for the professor to collect the signed forms in a blind fashion.

A couple of alternative situations may be present. If you are collecting survey data that is not part of the class (thus not student data), you could use electronic consent where no identifiers are used. Typically, they will click on an “I agree to participate” box or button which takes them to the survey without providing identifiers. This is an acceptable practice when the primary risk of the research would be a breach of confidentiality. If doing this, you should request a “waiver of documentation of consent” on the CUHSR application form. In cases of pre and post-data, another researcher, who is not the main professor, could match up pre and post-data and then de-identify data so that the professor is blinded. Qualtrics can also provide a unique code to a participant that they can use to take a post survey. The pre and post surveys are de-identified but can be matched by the unique codes.

SOTL Research Confidentiality

Privacy, confidentiality, and data protection are always concerns. Anonymity is the highest level of privacy, and this occurs when there is no link to any identifier to the data at any time in the research process. A survey sent electronically with no identifiers collected would be considered anonymous. Anonymity implies that the researcher is never able to identify the participants and any point. Confidentiality relates to the efforts by the researcher to protect the identity of participants. A good example of a confidentiality practice is to de-identify the data and replace it with a unique code early in the process. Re-identification can sometimes be an issue. This can happen when detailed demographics are collected in such a way that an individual can be identified. For example, suppose your data set contains information such as gender, race, class, zip code, and an activity, then the person could be identifiable by this data set (for example – an individual who identifies as a Hispanic female who is a sophomore volleyball player from zip code 61606 - based on this information the participant could be identified). Therefore, you should explain how each variable relates to your research, that is state how each demographic variable relates to your research question. Or you can state that your demographic data will be disassociated from the other research variables in the data set and demographics are only dealt with in aggregate to describe your participants. Describing your participants in this way is sometimes a requirement of publication. Typically, the data set should be de-identified in such a way that if a third party were to look at the data set, they could not see or formulate the identity of any participant. Detailed handling of the data confidentiality will need to be explained in the CUHSR application and in the consent form.

Most SOTL research is very low risk and likely will fall into an EXEMPT category. EXEMPT means that the research is exempt from the stringent federal regulations of human protection in research. Exempt research must fall squarely into one of the exempt categories and only CUHSR can make this determination. Risk will rise in SOTL research projects when a breach of confidentiality could harm the student’s reputation, financial standing, or employability. CUSHR will also consider research that may cause emotional distress or embarrassment. In these cases, extra steps will need to be taken to inform the individual of the level of risk directly and clearly in the informed consent and to then provide extra means of maintaining confidentiality or provide the means for the student to access help should they have an emotional reaction to the study. Studies with more risk will change the category status to EXPEDITED or FULL REVIEW status. In these cases, the bar to achieve approval is higher.

The Protection of Pupil Rights Amendment (PPRA) is another regulation that needs to be considered with doing research with minors (individuals under 18 years old). The PPRA specifies under most conditions the right of parents and guardians to inspect any instrument used to collect information concerning the following:

  • political affiliations or beliefs of the student or the student's parent
  • mental and psychological problems of the student or the student's family
  • sex behavior or attitudes
  • illegal, anti-social, self-incriminating, or demeaning behavior
  • critical appraisals of other individuals with whom respondents have close family relationships
  • legally recognized privileged or analogous relationships, such as those of lawyers, physicians, and ministers
  • religious practices, affiliations, or beliefs of the student or student’s parents
  • income other than as required by law to determine program eligibility

Parental/guardian consent is required before such instruments can be used to gather information from a minor. Typically, a researcher must be mindful of this provision and would normally get consent before an instrument would be used. The consent should state that the instrument is available for inspection. Also, if the research is being done in conjunction with a school, the school will likely have a policy related to the PPRA, and the researcher will need to comply with that policy. In cases where these topics are covered, a waiver of consent or documentation of consent cannot be granted by CUHSR.