Thursday, June 29, 2017

Was that email ethical?

I asked yesterday How would you respond to this email?

I read it carefully and did not reply.
The most striking thing was how generic it was. Although praising me and my work it never mentioned any specifics.
Sometimes when I get student inquiries I send them an email similar to that below.
Thanks for your interest.Please send me a copy of your CV and grade transcript.
I suggest you look at my blog under the label “hydrogen bonds” or "strongly correlated electrons" to get some idea of my current interests.
Also look at  “undergrads” and/or “Ph.D” to get some idea of my views and philosophy on supervision.
I suggest after looking at the blog you then write and send me two paragraphs:
one on why the science interests you and one on your perspective as to my philosophy.
After that, if you are still interested I suggest we then meet in person.
However, I did not send such an email for several reasons. I usually delete generic inquiries. If the student has not bothered to find out or articulate anything specific about me I doubt they are a very good researcher. I also want to work with people who want to work with me, not just anyone. Furthermore, tuesday was busy and I was not going to adjust my schedule for a student who was just showing up on a random day. 
So, I deleted the email.

I was fascinated that the next day I received a follow-up email.

Subject: Debrief Email: Follow-up from "Meeting: Prospective Doctoral Student"

Dear Professor McKenzie,

Yesterday, we, Associate Professor B... and Associate Professor M..., sent you an email inquiring about research opportunities as part of an experimental study on the effects of name identifiers on responsiveness to email communications. Our research is concerned with informal pathways to academic careers and involved random assignment of different sender names to test their effects on response rates from a large number of academics across Australia. Although the email was purportedly from a prospective research student, in reality this deceptive claim was a necessary element of the experimental design and the email was sent by us. We understand that this may cause concern. We are sending this email to reassure you that the data collected are anonymous, the study has the approval of the appropriate ethics committee, the deception was absolutely necessary for the integrity of the research, and only aggregate response patterns across groups, fields, and universities will be studied. 

Our study is titled “An Open Door? Experimental Measurement of Potential Bias in Informal Pathways to Academia.” All of this study’s data will be permanently anonymised, so there will be no identification of you or any individual with any response record. All names and email addresses will be permanently removed from the data and discarded. This study is only concerned with aggregate response rates. No individual response or lack thereof can indicate anything by itself, and your anonymised individual response will not be the subject of our analysis. 

Research involving humans in Australia is reviewed by an independent group of people called a Human Research Ethics Committee (HREC). The ethical aspects of this study have been approved by the HREC of the University of Sydney (protocol number 2015/757). As part of this process, we have agreed to carry out the study according to the National Statement on Ethical Conduct in Human Research (2007) (Updated May 2015). This statement has been developed to protect people who agree to take part in research studies. Because we are only measuring response rates to a single email, it was not possible to request consent prior to your participation. 

We greatly appreciate your participation in this study and will be happy to provide a copy of our report on the study results, once it is ready. We also understand that your response to our email took time and effort to write. We apologise for this. We believe the research we are doing is of considerable importance for improving understanding and informing policy. To receive our report, or if you have any other questions, comments, or concerns please contact  us ... [names and emails]

If you prefer, you may also or alternatively contact the Manager, Ethics Administration, University of Sydney. 

If you are concerned about the way this study is being conducted or you wish to make a complaint to someone independent from the study, please contact the university using the details outlined below. Please quote the study title and protocol number. 

Wow! 
I guess this email generated quite a firestorm because within 5 hours I received the following email.

Subject: Follow up research ethics

Dear Professor McKenzie,
Project Title: An Open Door? Experimental Measurement of Potential Bias in Informal Pathways to Academia

Project No: 2015/757

The University of Sydney has received a number of complaints in relation to this research project and its approval. The University takes these complaints seriously and the issues raised will be looked into by the HREC Executive committee. The project has currently been suspended.

All participants will be informed of the outcome of this review.

If you have further concerns about the way this study is being conducted or you wish to make a complaint to someone independent from the study, please contact the Manager of Ethics Administration using the details outlined below. Please quote the study title and protocol number. 


By the way, the name of my "student" was "Melindah Weelyrah". A Google search showed nothing. In contrast, one of my colleagues received an email from a name which he Googled and found a real web page at U. of Sydney for an apparently real person with relevant research physics interests. He did reply to the email.

Is this ethical? What do you think? Would you have filed a complaint?

Did the strong negative reaction from the "subjects" in the study stem from the "deception" or that they do not like having their time "wasted"?

This reminded me of the Brisbane bus driver racism study that caused a major controversy. The research results were overshadowed by the ethics questions and how the university treated the chief investigator.

9 comments:

  1. just copy and paste the project title in google
    you will get all. very sad to see this.

    ReplyDelete
  2. Hm, interesting indeed.
    I am not sure it is unethical; to gauge the response (of academics) as a fct of name (classes), one has to do something like this.
    Essentially the question is now whether asking hypothetical questions ("I'm xyz, can we meet") without saying they are hypothetical is ethical or not. I think it is not unethical.

    Regardless, the study is quite poorly designed, because from your description, your (non-) response had nothing to do with the name, but everything with the generic text in the email. This means the answers they would extract from their statistics are potentially flawed.

    Finally a somewhat critical question to your approach:
    I understand the reasons for your approach, but it does have severe consequences for "naive" students (see my comment on the original post). Aren't you expecting "perfect" students at the start of the PhD career despite the fact that that career is an educational one - i.e. they still have to learn things, even outside of physics?

    ReplyDelete
    Replies
    1. Thanks for another great comment.
      I think your criticism of my approach is fair. A naive student who is willing to learn is a good student. But finding those students requires a time investment and so is in tension with other demands.

      Delete
  3. I think these studies are valuable, but they should have used fictitious (ie, un-Googleable) names for all of the email inquiries. If they had done so, even with the generic text of the email, they likely would have been able to clearly distinguish response rates based on the name of the sender.

    The issue of implicit bias is real. The reason it is is "implicit" bias rather than "explicit" bias is that we tell ourselves a different story for our actions. The point is that Ross had a perfectly good reason not to respond to that email. But it is hard to know what implicit biases lie under that apparently rational behaviour. He could have responded anyway, saying that he could find a 10-minute hole in his schedule, if he wanted to. Maybe researchers would be more likely to make time in their schedules depending on the name.

    To be totally clear, I am not accusing Ross of any bias at all. Rather, I am defending studies like these as valuable and low-cost ways to expose, to us, the way that our brains and decision-making processes work, even and especially when we don't like the answer.

    ReplyDelete
    Replies
    1. Thanks for the comment.

      I agree that studies to show "implicit" bias are very valuable. The Brisbane bus driver study did this nicely. When I discussed it with some friends, I and a few others confessed that we would probably have acted in the same prejudicial manner as some of the bus drivers.

      I think the biggest problem with this study was the poor methodology, not the ethics.

      Personally, I do not believe that the ends ever justifies the means. It is too much of a slippery slope. Also, I think we should err on the side of caution when there are ethical ambiguities.

      Finally, we all need some perspective. This "act of deception" and "wasting some faculty time" needs to be compared to the kind of big scandals that ethics committees are meant to stop, such as the case of Henrietta Lacks.

      Of course, I am not saying that "small" ethical issues don't matter. Just, lets keep some perspective.

      Delete
    2. Ross,
      I think the discussion about "is this ethical or not" at this point results in yes or no answers.
      For the people saying "no", I would like to see *what* they find unethical. It's easy to say no, but there needs to be a spelling out of a concrete violation of some ethical code underlying that no.

      Unfortunately, the email from the oversight people did not mention any arguments put forward by the complainants.

      In my "I think yes, it's ethical" answer above, I gave an admittedly incomplete/poor effort to provide a reason ["hypotheticals are not unethical" combined with "saying something is hypothetical is unworkable for implicity studies" - though I admit that is an "ends justifying the means" argument].

      Finally I think any behavioral research into non-rational (instinctive/unconscious/...) behavior involves deception due to the necessity of avoiding making the subject conscious of the focus on these deeper level origins of behavior.
      If that is unethical, we can say goodbye to all of this research.
      That obviously leaves the unsolicited time-wasting argument still standing.

      And now I'll get of my soapbox.
      Interesting discussion!

      Delete
  4. I consider this sort of "study" as outright fraud, and grossly unethical.
    I have in fact at least once complained to the ethics office about such a troll.

    To be specific, I consider any academic study that interprets a "subject"'s response to a question as other than the literal direct response to a direct question as unethical.

    At least in the USA, one sees far too much of this. And it so frequently is tied into very very nasty politics that it should be prohibited at educational or governmental institutions.

    As a second, separate, concern I consider it far more worthy of my tax dollars to spend them on what appears today to be far-out hopes like "desperately searching for SUSY".

    ReplyDelete
  5. responding to Unknown at 5:34
    I think your definition would label most communication between (long-term) life-partners as unethical...

    To study the influence of boundary conditions, you have to create a focus and see what how boundary conditions affect the outcome. That is what is done here, and I do not think there is any other way.
    You essentially propose to ask "would you respond to name abc? And to name xyz?"
    That is obviously an impossible approach to measure implicit bias.

    ReplyDelete
  6. pcs 11:04

    .... they still have to learn things, even outside of physics?.

    This aspect is very important , if you are doing experimental work and you have to interact with other depts across the University. The student and the supervisor have to be very good communicators and have peoples skills. One supervisor who had phenomenal skills outside his subject was the legendary Max Perutz. Here is a sample of his greatness. An extract as by the legend himself below.

    "Making People Talk and Listen to Each Other

    Experience had taught me that laboratories often fail because their scientists never talk to each other. To stimulate the exchange of ideas, we built a canteen where people can chat at morning coffee, lunch and tea. It was managed for over twenty years by my wife, Gisela, who saw to it that the food was good and that it was a place where people would make friends. Scientific instruments were to be shared, rather than being jealously guarded as people's private property; this saved money and also forced people to talk to each other. When funds ran short during the building of the lab, I suggested that money could be saved by leaving all doors without locks to symbolise the absence of secrets.

    The above extract from the foll, web site.
    http://www.nobelprize.org/nobel_prizes/themes/medicine/perutz/index.html

    The last line leaving doors open in present times could be difficult. It is worth mentioning that 1962 , Crick and Watson both from Perutz group and Perutz himself won the Nobel Prize.

    There is an excellent review of the book
    Max Perutz and the secret of life, by Georgina Ferry in this web site.
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2222719/

    ReplyDelete