Ethics update: approval!

In my last post about this, I had just submitted my ethics application. In this post I’ll talk about the feedback I received, and how I have addressed those issues.

Administrative points

Some of the information that was required for the ethics proposal was missing. Rookie. This meant that I’d forgotten to put in some information about one of my supervisors. Easy fix.

I also needed to write about the purpose of the study for the prospective participants, on the Participant Information Sheets. I had written about this in too much detail (!) initially and so needed to make this more ‘user-friendly’. Essentially what this meant was re-wording the ‘purpose’ section to remove any jargon. Another relatively easy fix.

Opt-out vs opt-in methods of consent

My study has two main parts (which you can read about here – little has changed since this post). For the student part of the study (where I’m giving all students a questionnaire, and then conducting a focus group with a small number of pupils), I had decided I would use an opt-out method of consent. In the feedback, it was suggested that I needed to ‘reconsider this’. Fair enough; this is certainly a weaker form of consent than opt-in systems. To fix this, I needed to more precisely justify why I had chosen this method rather than opt-in.

My new-and-improved justification for maintaining opt-out included:

• An increased participation rate using opt-out, compared to opt-in (since parents/guardians are unlikely to actively opt their child in to participate, even though they don’t actually have a problem with participation).
• The participants (or in the case of students, their parents/guardians) are still getting the same information as they would otherwise, and is no more stressful for participants (Hunt, Shlomo & Addington-Hall, 2013)
• Students whose parents won’t opt-in are disproportionately likely to be from socioeconomically disadvantaged backgrounds. By using an opt-in method, therefore, a whole demographic (or likely a significant portion of it), are automatically being excluded, which reduces the trustworthiness and generalisability of the data (Henshall et al., 2021).

The next thing I needed to consider was how people would opt-out. For teachers before interviews, this is easy, they can just tell me they don’t fancy it. For over 550 pupils, however, this is more challenging. I initially had suggested parents could email with their child’s name, but actually this is a bit of a faff. Instead, they will now click on a link for a Microsoft Form and enter their child’s name there. Much easier.

Role of the co-investigators

Whilst I did originally acknowledge that there may be biases, I perhaps wasn’t clear about how I’d aim to acknowledge these in a way to make it clear what might have an effect on the analysis and conclusions. I don’t think it’s possible to remove bias, since humans inherently have values and ideals that will impact their view on things. However, acknowledging how those values and ideals have had an impact is probably the next best thing.

In particular, I hadn’t considered how I can make the most of the oodles of experience of my co-investigators (aka my supervisors). The first part of this means having regular discussions with them about decisions I’m making throughout the study; this acts as an opportunity for me to ensure I’m justified in making that decision, and that the reasoning is sound. In addition, I can get advice where needed. This will help to ensure more consistency in the way that the data is analysed, for example.

Details in thematic analysis

Finally, I needed to give more detail in how I would thematically analyse the qualitative data being collected. To be honest, I thought “I’ll work out how to do that when analysis time starts”, but with hindsight and the wisdom of this ethical feedback, it’s actually pretty important to know what I’ll do with data before I’ve collected it. I had just said “I’ll identify the themes that link aspects of the data together” but that’s a bit vague.

Before I continue, feel free to refer back to the post I wrote about my proposal, which includes some details about methods. In addition to that, remember I’m planning to use SRQ-A (questionnaire) data to choose students for the focus groups. I won’t write out all the details, but the rough process I’ll follow for the qualitative methods is here:

1. Code the verbal responses provided from the participants – what is the general meaning of their comment / response to the question or prompt?
2. By finding links between the data from different codes, I can identify any themes
3. For the student focus groups, I can also identify how the codes for specific students change over the course of the year
4. I can also use the students’ focus group responses to identify how they relate to the quantitative data provided in the SRQ-A.

In terms of getting to the level of detail needed around thematic analysis, Helen Kara’s book Research and Evaluation for Busy Students and Practitioners (2nd ed.) was really valuable. (Actually, it’s been valuable for other things too). I don’t have the book on me at the moment but will edit this post with a proper reference later.

What’re my next steps?

The most pressing thing I need to work on is running a pilot study in the next few weeks. This basically just means making sure the data generation tools are actually ready to be tested with a small sample of students. I also want to be confident that I can conduct a focus group properly.

Concurrently, I’m starting to make notes on the methods I’m using in a skeleton thesis document (!!!!)

References

Henshall C, Potts J, Walker S, et al. (2021). Informing National Health Service patients about participation in clinical research: A comparison of opt-in and opt-out approaches across the United Kingdom. Australian & New Zealand Journal of Psychiatry. 55(4): 400-408. doi: 10.1177/0004867420973261

Hunt, K. J., Shlomo, N., and Addington-Hall, J. (2013). “Participant Recruitment in Sensitive Surveys: A Comparative Trial of ‘opt In’ versus ‘opt Out’ Approaches.” BMC Medical Research Methodology, 13(3). doi: 10.1186/1471-2288-13-3