When researchers toil in the toilets
A topic in 'Science & Ethics'
We usually consider that people should only participate in research studies if they have given voluntary, informed, consent. Certainly, if they are going to be required to give up considerable time, or open up about issues that may be sensitive or personal.
Read about voluntary informed consent
In many studies, the researchers know the identities of participants, but promise them anonymity in any public reports. (Offering anonymity is the default in many areas of research – although there may occasionally be circumstances where that is not appropriate.) But some observational studies may simply watch the behavior of unidentified individuals. Such observational studies reported on anonymous people without any information that would allow their identities being inferred – and even the researchers do not know who they were.
What if a researcher were observing people arriving in a town by train, to see what proportion continued on foot, compared with those who got on a bus, or perhaps had a parked car waiting, or cued for a taxi? Would any harm be done if observation was covert and no one observed knew what was going on?
Is research that observes people covertly, without their knowledge or consent, always unethical? Might there even be a need sometimes to avoid people knowing their behaviour is being observed?
Now, please wash your hands…
Consider a study carried out in a public restroom (women's toilets) to see what proportion of users washed their hands before leaving the room.
Would it be acceptable to make these observations secretly, without seeking consent of those being observed? Would covert observation (say a secret camera) be unethical?
Here is a brief extract from a real research study:
"Women were observed after elimination in a public restroom to determine the effects of the presence or absence of an observer upon the likelihood of handwashing. 18 of 20 subjects in the presence of another and 3 of 19 who were not observed washed. The difference in frequency of handwashing between the two groups was clearly significant."
Pedersen, Keithly & Brady, 1986, abstract
So, in this particular study, two conditions were used.
In one condition, those using the facilities were aware someone else was present to observe their behaviour. In that condition nearly all the 'subjects' washed their hands after using the toilet.
In the other condition, the 'subjects' were not aware they were being covertly observed. In that condition, very few washed their hands after using the toilet.
The contrast is stark: being aware of being observed seems to change behaviour for many people. (These are quite small samples, from one study undertaken in one place – but the shift between 90% hand-washing when someone else was obviously present, and only about 16% hand-washing when apparently not being observed is…shocking.)
In this case, relying on a situation where people know they are observed, and so can choose to opt out of the situation or change their behaviour means that research conclusions (e.g., nearly everyone washed their hands after using the toilet) could be distorted because of the fact of the research itself. We do not only do research to find out what happens when we do research but because we hope it also gives us insight into what normally happens!
As I asked above:
Is research that observes people covertly, without their knowledge or consent, always unethical? Might there even be a need sometimes to avoid people knowing their behaviour is being observed?
One consideration here is that no one who is observed in research of this kind can be identified. No details are provided that allow us to infer whether Julie Smith, of Tucson, Arizona, 31 years old, dental nurse and mother to three, divorced from Hank, washed her hands or not!
Clearly research that allows participants to be identified raises additional ethical questions. Generally in research we look to avoid participants' identifies being determined: although there may be circumstances where that is not viable, or would need redaction of important contextual information – and there may be some circumstances where identifying participants is appropriate.
Read about anonymity of research participants
The study of hand-washing discussed above is somewhat different from my hypothetical survey of the ongoing transport options of train travellers. The railway station is a public place where we should expect others to be able to observe us.
What about a public restroom? We might expect a higher level of privacy – as surely there is a common understanding that using the toilet is a more private matter than, say, getting off a train, or having a meal at a restaurant. (That is there are cultural expectations about privacy relating to shared cultural taboos.) Of course, in the published study all that was observed was hand-washing behaviour – it is not as if anyone tried to publish research on observations of anyone actually using the toilets.
No one would do that...surely?
Personal Space Invasions in the Lavatory
Now, consider another study in which 'subjects' were not even aware they were being observed for the research (so, surely, were not harmed in any way?) and are in no way identifiable in the published report.
In this study covert observation was again used, and (like the hand-washing case) it does not seem feasible the work would have been possible had volunteers had to give consent.
Under these circumstances, is covert observation ethically justified?
You may feel that-otherwise questionable-research activities (here covert observation) could be justified if the research question was important enough that the research might lead to useful outcomes.
(This is also discussed in the account of 'An experiment on memory?')
In the study considered here:
"The hypotheses that decreases in interpersonal distance would lead to increases in the delay of micturation and decreases in the persistence of micturation were tested'
Middlemist, et al., 1976, p.544
In other words, the researchers suspected that if, when someone was emptying their bladder, a stranger was very close by, then this would have two effects. Firstly, if would take the person longer to begin urinating, and secondly that urine would be excreted for a shorter period of time.
These probably seem reasonable hypotheses as most people (a few exhibitionists excepted) would probably agree that it is easier to urinate in private than when you have someone standing very close by you. This effect likely has an evolutionary basis – to some extent a person is more vulnerable when urinating, so it make sense that we might be stressed in having to do so standing very close to someone else.
Research is undertaken to develop knowledge. Knowledge may be seen as simply answering interesting questions (that is, questions someone finds interesting) or because the information gained has useful applications. (Often 'pure' research is later found to have applications which were not foreseen at the time – but much planned research is justified to funders as having specific likely applications.)
To what extent, if any, do the likely applications of research knowledge impact on the ethical worth of research? Does this depend on the 'costs' of the research (to the researchers and funders? to those participating?)
We might take a view that knowledge is always a 'good' worth acquiring: but that like all things we might want to acquire, it is not worth having regardless of costs.
Spoiler alert
At the risk of jumping ahead (as I think it is the methodology which is most relevant to the ethics here), the researchers found their hypotheses were supported by their data:
"Micturation delay increased from a mean of 4.9 seconds in the control condition to 6.2 seconds in the moderate distance condition to 8.4 seconds in the close distance condition.
Micturation persistence also showed significant differences … from 24.8 seconds in the control condition to 23.4 seconds in the moderate distance to 17.4 seconds in the close distance condition"
Middlemist, et al., 1976, p.455
That is, the observed men started urinating quicker, and continued for longer, when they did not have another man standing close by them. (Well, who would have guessed?)
How did they know that?
Now you might wish to reflect a moment on how this data was obtained – on what actual observations would be needed to collect this kind of information, and on how those observations might be achieved in a covert manner.
Do you think there is an ethical way to covertly collect data on the timings of men's micturation?
The research paper explains how the methodology was developed for the study. Initially a naturalistic approach was taken-that is like a field study where the researcher does not seek to intervene, just to observe:
Read about naturalistic research
"An observer was stationed at the sink facilities and appeared to be grooming himself. When a potential subject entered the room and walked to a urinal, the observer recorded the selected urinal and the placement of the next nearest user.
He … recorded the micturation delay … and the micturation persistence…
The onset and cessation of micturation were signaled by the sound of the stream of urine striking the water in the urinal."
p.543
So, the researchers' assumption what that they would observe a range of behaviours – sometime a man would seek to use a urinal next to one already being used, and sometimes they would approach one removed, and sometimes they would seek to be as far from other users as possible. This would provide three conditions:
- control (no other user)
- moderate distance (another user, but not immediately adjacent)
- close distance (another user immediately adjacent)
So, although the researchers conjectured that men would find urination easier when their 'personal space' was not being 'invaded', they nonetheless assumed that observation would provide instances of their different conditions. (How realistic was that assumption?) However,
"Of the 48 subjects recorded, none selected a urinal immediately adjacent to another user…[So instead] Men who entered a three-urinal lavatory to urinate were forced to use the leftmost urinal. A confederate was placed immediately adjacent to the subject, one urinal removed, or was absent from the lavatory. An observer stationed in a toilet stall timed the delay and persistence of micturation."
p.543
So, now the research design had shifted from naturalistic observation of how men normally behaved when emptying their bladders to an intervention study where men were manoeuvred into using particular a particular urinal. A confederate was used – that is an associate of the researchers playing the role of someone just using the facilities.
Does this effect the ethical status of the research? Does the study become any less ethical when users of the toilets are being manipulated in this way? (If so, why?)
When the study was piloted (tested out) this new procedure worked well enough. By the main study took place in a different restroom, leading to a new complication. 'Observation' may bring to mind watching but here, as mentioned in a quote above, the observation used hearing – and was based on timings of "the sound of the stream of urine striking the water in the urinal". Unfortunately this method failed.
"An observer was stationed in the toilet stall immediately adjacent to the subjects' urinal. During pilot tests of these procedures it became clear that auditory cues could not be used to signal the initiation and cessation of micturation. The urinals were so silent that even the confederate standing adjacent to the subject could not hear the urine striking the urinal."
p.544
Perhaps put yourself in the position of the researchers.
- What would be the appropriate response to this hitch?
- Abandon the study?
- Move the research site to a restroom where the observation method was reliable?
- Find a different way to make the necessary observations in the existing research site?
Modifying research design
The researchers chose to find an alternative way to make observations:
"Instead, visual cues were used. The observer used a periscopic prism imbedded in a stack of books lying on the floor of the toilet stall. An 11-inch (28-cm) space between the floor and the wall of the toilet stall provided a view, through the periscope, of the user's lower torso and made possible direct visual sightings of the stream of urine.
p.544
The observer, however, was unable to see a subject's face"
Reassuringly, the researchers did not look at the men's faces, but only at 'lower torsos' to watch the men urinating.
Presumably, when this research was carried out (in the 1970s), it was considered quite acceptable-at least it must have been by the reviewers and editor of the Journal of Personality and Social Psychology-but I wonder how many of us would find it perfectly acceptable to be spied on when using a toilet by an unseen stranger using a periscope. Indeed, perhaps there is a case here for suggesting a criminal act was committed?
Do you consider this research ethical? How do you justify your position? Has there been any harm to those observed? If so-How were they harmed? If not-Why should researchers not collect data in such ways if those observed remained unaware and no one knows who they were?
Finally, imagine the same study had (as far as it transfers) been carried out not in a public restroom, but in a farmyard to test the "hypotheses that decreases in interpersonal distance [between pigs] would lead to increases in the delay of micturation and decreases in the persistence of micturation were tested".
Would the subjects being pigs rather than humans make any difference to your judgement of the ethical status of the study? WHY?
One of the other vignettes explores the use of animals in scientific research.
One interesting feature of this study is it shows how initial research designs may prove to be problematic, and so researchers may have to make changes to their plans to collect their data.
That is a methodological issue, but perhaps also impacts on how researchers consider the ethical status of their work.
The original design required people standing near the washbasins in a men's toilet (where they could be seen by users of the toilets, even if they were pretending to comb their hair rather than recording data), noting which urinals were being used and making timings by listening.
This shifted to having confederates pretending to use urinals to ensure where subjects would stand, while they were bring spied on by an unseen researcher using a periscope.
Is it possible that the ethical acceptability of the study shifted without any due consideration being given to this aspect of modifying research design?
(Read about a medical study where researchers continued their {already highly questionable} research despite highly pertinent changes in external circumstances which impacted the ethical status of the study.)
Work cited:
- Middlemist, D., Knowles, E. S., & Matter, C. F. (1976). Personal Space Invasions in the Lavatory: Suggestive Evidence for Arousal. Journal of Personality and Social Psychology 33(5), 541-546.
- Pedersen, D. M., Keithly, S., & Brady, K. (1986). Effects of an observer on conformity to handwashing norm. Perceptual and Motor Skills, 62(1), 169-170. doi: 10.2466/pms.1986.62.1.169
