In this post, I will provide step-by-step instructions on how to prepare and submit an IRB (Institutional Review Board) application. I am currently at University of Oklahoma, so I will use their system to illustrate the process. However, I am sure it will apply to other institutions, because all IRBs abide by the same Federal regulations and APA guidelines. I have also successfully submitted IRB applications at Humboldt State University, where I completed my Master’s degree (about my research).

I have submitted dozens of IRBs. Each one has at least one modification. One has had 13 modifications approved. There are benefits to modifying an existing IRB rather than creating a new application. I will cover that after explaining the initial process. The point is that it is not difficult to get your research approved by your IRB. And yet… I have been asked time and again to help people with the process. As usual, I am writing a blog that I can point people to next time I am asked to explain something.

Is it really more difficult than I am letting on?

Not really. But the truth is, they don’t make it easy. Oh, your IRB office (OU’s website) will try to help you… I promise!! One of the best things you can do is pick up the phone and call your IRB office when you have a question. You will talk with the office staff. They are not the actual IRB (the board of people who review your proposed research). Rather, the staff is in charge of the basics. They vet the applications for required components before sending to a Board member. They ensure that everyone’s training (CITI) is in order. The staff probably know the answer to your questions. If they don’t, they can help you figure out who does.

Below I will outline the general process. I will include advice on how to deal with certain issues. If I miss something you came here wanting to learn about, let me know and I will add it.

The quick list:

1. Complete CITI training if you are not up to date.
2. Prepare all your documents (consent, stimuli & materials, surveys, recruitment) beforehand.
   • OU Forms and Templates
   • It helps to have pre-registered your study first. It will help clarify your thinking and make your work easier in the long run.
     • You can prepare a document with all your materials that will serve for pre-registration and to meet IRB requirements.
3. Go to your institution’s IRB app site. (OU’s is here.)
4. Read carefully.
5. Follow the instructions.
6. Answer clearly. It really helps not to make grammatical and/or spelling errors.
7. Don’t be wordy. Nobody wants to read more than is absolutely necessary, and more words are just more opportunities for you to put your foot in your mouth.
   1. It’s always better to be told to add something (they will tell you how) than to include something that (accidentally and unnecessarily) rings a lot of bells.
8. Whenever possible, give yourself wiggle room.
   • Describe the study as broadly as possible.
   • For example, rather than stating “Participants will be randomly assigned to read Story A (fiction) or Story B (nonfiction)”… Say “Participants will be randomly assigned to read either fiction (e.g., Story A) or nonfiction (Story B). I usually give 2-4 examples of stimuli per condition.
   • Similarly, don’t say “We will use this recruitment wording: (‘blah blah blah’).” Rather make heading “Sample recruitment scripts” and list 4-5 samples.
9. Read carefully, follow instructions.
   • Almost every IRB fiasco I’ve heard of resulted directly from the failure to read carefully and follow instructions.

Pre-PS: If this information has helped you at all, do not hesitate to buy me a coffee!

A few suggestions for the pre-application work

If using an online survey, you will need to prepare it fully before completing your IRB application. I have Qualtrics posts already that cover Embedded Data, Random Assignment, and Using Logic. More will come.

Here I will suggest you refer to your institution’s IRB site for templates for consent, recruitment, and other forms. You will want to download a copy of your online survey, if applicable (print to pdf). I suggest keeping a document with all of your materials (the same one you upload to osf if you pre-register). For scales, write out all the items on separate lines. Then you can copy-paste into Qualtrics. Use the Word list function to add numbers if you want (easier to get rid of them if needed).

Check with your PI (if you aren’t the PI) for any lab-specific procedures. For example, your PI may want all consent forms to contain certain wording. The lab may already have templates or examples of all the material you will use.

Step-by-step for the IRB application

After you sign into your IRB site, you will see the options below. Add a new study. (When you have any task open, it will appear below those buttons. Possible tasks include not-yet-submitted applications and modifications, or forms that have been returned to you for corrections, etc.)

When you first start a study, you will have to choose the campus/location. I shouldn’t need to tell you how to do this.

Titles: Step 1, section 1.0

This page seems trivial, but there are two things to take into account. First, when choosing a title, remember that it must appear on your Consent form or information sheet. A few participants might actually read your consent/info sheet, so don’t choose anything that you don’t want them to see. Second, unless you will be the only person on the study, spell everything correctly. If you misspell anything, please don’t put me on the study. It will bother me every single time I notice it on my “My Studies” page.

After the title, you will set up departmental access in section 2. I’ve only ever had the option to choose psychology… Presumably people belonging to other departments will have the option to choose their department.

Section 3.0: Key Study Personnel (KSP)

You are going to get up close and personal with this page. First, you’ll need to set it up correctly. Then, if you run studies across multiple semesters with changing lab personnel, you will have to update it.

You will add the Principal Investigator first. That will be your mentor or faculty advisor.

The IRB application form changes all the time. Currently, we can add a co-principal investigator (one), but that was not always the case. We can currently also add as many sub-investigators as we want, as well as research support staff. These will be fellow grad students, undergraduate research assistants, or staff involved with the project.

Under the study contact area, you will need to add your PI (unless you are the PI, and if you are, you probably don’t need to be reading this). If you are a student, add your faculty advisor.

Note: Everyone on the study must have completed CITI training and be in the IRB system. Sometimes, for mysterious reasons, there will be a person whom the system cannot find. You can add them in a note (at the end of the Study app) and upload a copy of their CITI training certificate. It’s also possible that someone has a different name in the system. Perhaps they have got married. Perhaps they use a different name. I try to remember to ask RAs about this at the first lab meeting they attend.

The next two sections (4 and 5) are self-explanatory.

Section 6: Participant information.

In the next section, you will add a row for each group/population you will be sampling. If you are only using one (e.g., the departmental research participation pool), you will only have one row.

I’ve started with a row for college undergrads from the research pool (SONA). You will need to specify the number of cases for each group. Your consent form must include the total number of participants for the entire study. That is, if you plan to recruit 400 undergrads and 400 adults on social networking sites, then you consent form must state that 800 people are expected to complete the study. You will need to increase this number if you end up running more participants.

Make sure you complete the entire row (age, race/ethnicity, language).

I usually put the age range from 18 to 100… not that I’ve ever had a centenarian. My oldest participant ever was 85 though. For studies using the departmental pool, you might want to include 16-17 year old students. They can (and do) get parental consent to participate, and they will complete your studies, even if you state clearly that participants must be 18. This won’t complicate your IRB process (as long as they aren’t the focus of the study), just make a note of your rationale for including minors. Targeting minors and/or race/ethnicity will make your study subject to full or expedited review.

You might have to scroll to the right to see everything. At the far right is a question about language. If not everyone speaks English, you will have to prepare translations.

Other participant groups (MTurk, general population adults…)

For adults recruited from the Web, I usually say something along the lines of “adults recruited on social networking sites.” You could also say “via email” or “in the mall.”

Amazon.com‘s Mechanical Turk can be a good source of paid participants (see my post on the topic). Be careful to vet your data carefully though! I’ve just started a Facebook group where researchers can talk about the pros and cons and pitfalls of using MTurk samples: Academic MTurk Users. Feel free to join!

As far as the Study application goes, IRBs nowadays are familiar with MTurk. If you include worker qualifications (such as an age range), make sure that is reflected in your IRB app:

Section 7: Research Design.

Here you will state your study design in plain English. If you have submitted a pre-registration, you can usually just copy-paste your Study Description from the osf form (if it’s short and to the point). You will need to update this with each substantial modification (more in a separate post). You will include such things as whether it’s cross-sectional/correlational or if participants will be randomly assigned to experimental vs. control conditions…

This used to be part of a Protocol that you could upload. I preferred the old protocol template, because it served as a precise record. Here, you can just put “survey instrument.” Of course, you will have to upload a copy of your survey, but it’s much better to list all the questionnaires and tasks that are in the survey. I do this in a separate “materials” document, but that needn’t be uploaded to the IRB site.

Again, be succinct. Do not overshare. It is better to be asked to add things than to be told that under no circumstances can you do that (because it looks suspicious). Remember, the IRB is going to have access to ALL your material. They will click through your surveys. You will upload any in-lab materials. They can arrive at their own conclusions.

Don’t hesitate to ask for advice.

If you are in doubt about the suitability of some stimulus, pick up the phone and call the office. The staff will tell you if you can do something; they will also tell you how to present it in your app in a way that is not going to raise red flags.

A phone call can save you hours of work and stress. (If you are at OU, the staff are all really nice!) Of course, you should already have talked with your advisor about these things. On the other hand, s/he doesn’t want to have to hold your hand the entire time.

Section 8: Study sites.

This is fairly self-explanatory. See image below for online example. You might also say in the lab, or in front of a movie theater…

Section 9: Key Study Personnel (KSP)

Everyone whom you added in Section 3 needs to be added here too. Fortunately, you can copy rows once you have created them. In 9.2, you will put something along the lines of “The two PIs designed the study together and meet on a weekly basis. The research staff has been trained to run the study and attends lab meetings every two weeks where everyone is briefed on changes and challenges.”

Section 10: Risks and benefits.

This section is important, because it demonstrates that you have thought through your study and are aware of any risks. Mostly, in psychology, these will be emotional. In fact, mostly they will be nonexistent. Sometimes neither you (the research team) nor the IRB will notice things that some participant will make a fuss about. We had the Dead Chicken incident a few years ago…

You can avoid issues here by choosing stimuli that is not going to cause concerns for the IRB. For example, if you use TV shows as stimuli (we do), you can avoid using scenes with explicit content. Yes, most undergrads are seeing far worse, but that’s on their own time.

Even with minimal risk, you will need to mention it. Besides the Dead Chicken incident, we’ve had other potential emotional triggers. Last semester, participants were randomly assigned to view trailers of films or news about refugees. My IRB application was returned with the stipulation that I add the language in the image below:

One way to deal with potential objections is to give participants an explicit way out. We are currently running a study in which participants may be asked to get into the mind of a serial killer. Because some people might not want to do this, we give them the option to choose another task.

Section 11: Recruitment

Just list how you plan to get participants. I specify the procedure for each group (see image below). Your eventual paper will have a participants section in which you can use very similar language. If you’ve done a proposal, you can judiciously copy-paste from that here.

Section 12: Compensation

Again, make it easy for the IRB to read. Use lists, tick the correct options.

Section 13: Informed Consent

The important part about this section is minimizing coercion.

If you have met certain conditions, you will get the Waiver of Signed Consent section (online..)

Section 15: Funding

Sadly, I’ve never had to check anything exciting here, as all of my research has been internally funded. This includes any departmental funds and research grants obtained through the university.

Section 16: Privacy and Confidentiality

Just make sure you tick all the appropriate boxes. Unless you are taking down names, addresses, phone numbers and the like, you won’t have identifying information. If you are having participants sign hard copy consent forms in the lab, you will. However, these forms need not be associated with the data.

What if I could somehow associate my data with the identifies of my participants?

Don’t do it. If it would take you five steps and several merged files that you don’t even have access to (e.g., SONA), the IRB doesn’t need to know about it. Avoid the possibility by never storing files with sensitive data (names, IDs). You shouldn’t be doing that in any case without taking appropriate precautions (previously stipulated in an IRB form). Such precautions include storage in secure locations (whether physical or virtual). Secure means locked, limited access, password protected, and so forth.

What about matching participants in longitudinal data?

Whether the data are entirely physical (e.g., hard copies) or entirely digital, you will need to assign participants id numbers and match them to any identifying information in a separate file. The keyed file should be kept in a separate location from the data file.

If you are using departmental prescreening and need to match data with a later survey (in lab or online), you will need to match and merge datasets. For online data, you can match with the SONA id (if you set up the survey flow to collect it). In some departments, students will know their SONA id, so you can ask them to input it. This is great, because the SONA id is not sensitive data. Unfortunately, OU psych students do not know their SONA id. You can use their student ID or initials to match data. I use initials plus dates to match. The best thing is to keep a “matching” file, where you have the SONA id and whatever you use to match.

If you are sent the prescreening data with sensitive information, delete everything you don’t need immediately. I also delete the file from my email account. Make a separate matching file if you have sensitive information (e.g., are matching with student IDs).

What about MTurk?

If you are collecting worker IDs with MTurk data (I do, to facilitate payment to deserving workers), then potentially you could match the data to a real person… That is, you could do so if the person contacted you and revealed their name. On their own, worker IDs tell you nothing.

Here’s a hint: if a few Workers email you, resist the urge to go through the painstaking process needed to connect their data with the identifying information you now have. In other words, this really isn’t an issue. I have only once described the process of MTurk on an IRB application in enough detail that mention of Worker ID happened. It wasn’t necessary. Unless you are very unlucky (or cause red flags to go up), you won’t need to address this.

Data storage.

It used to be appropriate to state that data would be kept for five years after publication as per APA guidelines. Now, you can simply refer to best practices in Open Science and state that you will make data available indefinitely. If you have identifying data and plan to keep it, then you will need to specify how this is going to be done securely.

Section 17: Application type

Think positively, and select the least onerous level of review you think your study will need! I always select exempt (with one exception when we planned a developmental study. Ended up a lot of work for nothing, since we never carried it out.) This doesn’t mean it’s actually exempt from anyone ever laying eyes on it. It just means that staff will look carefully at it to determine what level it actually merits, and a single IRB board member will verify and sign off.

Sections 18 and 19 are Conflict of Interest and HIPAA. They should be self-explanatory. Your institution will have policies regarding COI (read them if in doubt, or call your IRB office). HIPAA (Health Insurance Portability and Accountability Act) is all about protecting patient privacy. You should know if you need to worry about it.

Study app section 20: The end!!

Not really, because you still have to upload documents… but this is where your faculty advisor will tick some boxes that assures awareness and compliance. If you are the PI you can do it yourself. If not, after document upload you will click on “Notify PI” and the system will send him or her an email.

Document uploads

First you will upload your consent form(s). You will probably need one for each group of participants (e.g., undergrad pool, Web, MTurk). Don’t forget to select the language it’s in.

In Part 3 you will upload everything else. This will include recruitment material and data collection instruments (pdfs of the survey exported from the survey platform [e.g., Qualtrics] or digital copies of any paper forms, etc.). It may include letters of support from school administration, funding information, and so forth.

The end!! (until you get it returned with stipulations)

What if it comes back with stipulations?

Unless you are a lot better at this than I am, you will have the form returned with a list of things you need to fix. The first round may simply be errors that the staff catches. Later, an IRB board member (or staff, if exempt) will return it with the status of “Approval pending…” a series of stipulations. These are just things you need to change if you want it to be approved.

My record is 19. Usually there are a few things I need to change and/or clarify. It’s not a big deal, just do what you are told. Call your IRB office if in doubt. Consult with your advisor or mentor.

Once your application is approved, you may start to run participants.

If you change anything at all about your procedure, increase the number of participants,change KSP, or add a subject pool, you will need to modify the IRB. I will cover modification in a later post. In general, it’s a lot easier to modify an existing IRB application than to change an old one. This is possible as long as the new study is within the scope of the original one (ie addresses the same research questions).

However, you want to keep your records easy to manage. If you cannot remember the first study you ran under this IRB, it’s time to start a new application.

Good luck and let me know if I’ve missed anything!