|LETTER TO THE EDITOR
|Year : 2022 | Volume
| Issue : 2 | Page : 217-218
Online survey as a data collection tool – The good, the bad, and the ethics
Fayiqa Ahamed Bahkir
Department of Ophthalmology, Karpaga Vinayaga Institute of Medical Sciences and Research, Chengalpet, Tamil Nadu, India
|Date of Submission||23-Nov-2020|
|Date of Acceptance||02-Jun-2021|
|Date of Web Publication||30-Jun-2022|
Fayiqa Ahamed Bahkir
Department of Ophthalmology, Karpaga Vinayaga Institute of Medical Sciences and Research, Chengalpet, Tamil Nadu
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Bahkir FA. Online survey as a data collection tool – The good, the bad, and the ethics. TNOA J Ophthalmic Sci Res 2022;60:217-8
The concept of publication ethics, be it data collection, analysis, or publishing, is vast and forever evolving. With the emergence of the COVID-19 pandemic, most studies requiring patient handling have been suspended. The trend of data collection has shifted to an online platform, which is currently the most economical and practical way to get the data collection tool to and from the study population.
But with newer avenues of research opening, the minor flaws in the methods have come to light.
An age-old question that has come up time and again is that of ethical clearance for a study conducted online. Researchers have questioned if any single Institutional Human Ethics Committee can give the clearance for a study that is to be conducted online, which broadly means that the data can come from anyone in the world, depending on what kind of target population parameters the study has.
There have been a few instances where the need for creating an independent, autonomous body for ethical clearance of studies has been called for. This organization would then handle proposed research work falling under two domains: studies that do all their data collection online, and studies conducted by independent researchers who are not a part of any institution.
Currently, conducting a study as an individual who is not affiliated to an institution is a Herculean task, considering the fact there is no independent body that gives ethical clearance for a study to be conducted. A person needs to associate themselves with an institution to conduct any kind of research work (this scenario is true for most countries till date).
When it comes to data collection on an online survey, the reach of the survey is limited to the resources of the institution or individual conducting the study. Access to mailing lists, which is how the survey is sent out, is usually restricted. Even if access is granted, the mails are usually filtered and sent to the spam folder and never reach the intended recipient. Another option is to send them out to the people you know, hoping that they will send it to people on their lists, and so on. This method is largely dependent on others propagating your study.
In recent times, advertisement of a survey on certain websites has been done to increase the reach of the data collection tool. The type of website and people accessing it largely influence the responses in the data. Moreover, these advertisements are paid, and it again boils down to the fact that the person with more resources can get his study a wider reach. Yet another method of online data collection is to give your data collection tool to commercial survey-taking sites that will guarantee a certain number of responses from your target population, provided you are willing to pay a hefty amount of money per filled questionnaire.
Certain survey-taking sites do exist where students upload their theses and wait for responses from other students who have their theses up on the site. Although this is a very effective tool to reach the target demographic if the study aims at a particular age group, there are instances where, just to finish a number of surveys in a short time, the responses given by the survey-taker are absent-minded. These sites usually work on a rank basis, wherein the more surveys you complete, the higher rank your survey gets, which increases visibility, and therefore, more responses. In an attempt to gain a higher rank, one might be tempted to complete as many surveys as possible, while not even paying attention to some of the questions, and therefore, rendering the completed survey quite ineffective for the study being conducted.
In these pandemic times, since patient contact is limited, most of the studies done are awareness and prevalence studies. When this becomes an online model, awareness tends to be overestimated, simply because the participants consider the data collection tool a measure of their knowledge, so they tend to try and “perform well” in most cases.
When prevalence studies are conducted in the clinic, and a patient is diagnosed as high risk for a condition, say metabolic syndrome, treatment or lifestyle modification measures are usually initiated in the same sitting as that of the diagnosis. In case of an online data collection tool, this can be delayed; and in cases of anonymous surveys, the patients might not even know they have been categorized as high risk for a medical condition.
A few tips for a hassle-free online survey
Despite its minor flaws, we must consider online data collection a much-needed silver lining in the dark cloud of the novel coronavirus pandemic. This method has a very quick reach and makes data collection almost effortless once the initial questionnaire has been prepared. A few ways do exist by which to make the process simpler, and ethically sound.
- In case of any study, there exists a need for patient consent. Although the action of clicking on and participating in an online survey implies consent, it is always better to put up a separate page for informed consent before your questionnaire is displayed.
- People tend to be cautious about giving out their names and age on an online platform. Most surveys do not ask for names, but they do need an age to be typed in. Considering most people refrain from filling up this question, asking for an age range in multiple-choice format is better than asking for an exact number.
- Every study participant is a volunteer and should be given the right to withdraw from the study at any time. This should be clearly mentioned, and the participant can be asked to enter a unique ID in a mixture of numbers and letters (e.g. initials of their name with their birth date and year), so that if they want to withdraw from the study in future, they can simply tell you their code, and their information can be deleted.
- People, in general, do not like typing out answers, a multiple-choice format is always preferred. If a question allows more than one choice, this should be mentioned clearly in the question in parenthesis.
- In case of unavoidable long surveys, where the attention of your target population might slip, and they might choose any option just for the sake of filling it up, a couple of attention checkers can be included. This consists of questions such as “Please choose the third option in this question,” and if they do not do so, their response is discarded.
- Adaptive questioning can be used wherever possible. For example, if your exclusion criterion is the smoking population, a question can be asked “Are you a smoker?” with options Yes and No. If they click on “Yes,” their response is recorded, and they are taken to the end of the survey. Only if the chosen option is “No,” they are taken to the next part of your study. This prevents unnecessary waste of time on both sides.
- Always try and beta-test your questionnaire before fielding it. Once an online survey is up and running, any flaws will be discovered only after the data have all been compiled and analysis has begun. It is always better to have a group of people beta-test your survey and give you feedback, which will not only allow you to avoid a disaster later, but will also ensure a higher number of completed questionnaires being submitted.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.