University of Texas at Austin
Skip to page content.
Information Technology Services
About ITS ITS Services Contact ITS ITS Departments ITS Employment ITS News ITS Help Desk

| | | | |
| header |

Ethical Issues and "Netiquette"



Advantages of Online Surveys

Disadvantages of Online Surveys

Designing the Questionnaire

Creating the Questionnaire

Data Collection

Data Analysis

Getting Good Data

Ethical Issues and "Netiquette"

Putting it all Together



Human Subjects Approval at UT

Before beginning your study, your methods must be reviewed by an ethical review board. At The University of Texas at Austin, any original academic research involving human subjects must be approved by UT’s Institutional Review Board. UT staff who are conducting administrative research are not bound by the same guidelines; however, the ethical guidelines for academic research provide a good standard which administrative researchers should be aware of. For more information about policies governing research with human participants at UT, see the Institutional Review Board Policies and Procedures.

As you would with any other research, you should provide respondents with a description of the online study, the possible risks to the respondent, a privacy statement, and other important information. Because some respondents may be unfamiliar with the computer-based methods you will use in your study, you should explain in clear detail exactly the procedures that will be used to track their responses and preserve their confidentiality.

You should also be aware that the Institutional Review Board may have special requirements for your online survey, particularly those that deal with vulnerable populations or sensitive topics. Please contact the Office of Research Support and Compliance for more information.

Smith & Leigh (1997) recommend that online researchers achieve informed consent in the following manner:

  • After reading the informed-consent information on the survey Web page, subjects enter their e-mail address.
  • The host server for the survey automatically generates a unique password and sends it to the subject's e-mail address.
  • The subject enters the password in order to provide their consent and access the survey (the informed consent form includes the information that, by using this password, they are giving informed consent and that they are capable of giving informed consent).

If the Institutional Review Board requires your online study to include informed consent, you should contact the TeamWeb consulting desk for more information about how to program this type of application.

Skipping Questions and Withdrawing

In order to reduce the possibility of psychological harm to subjects, UT ethical guidelines require that researchers allow subjects to skip questions or withdraw from the study. Allowing subjects to skip questions or withdraw also guarantees higher-quality data for the researcher, because bored or upset respondents may complete your survey with false replies.

Skipping questions. On the Internet, it is easy for respondents to accidentally skip a question. For example, a respondent may click a response with their mouse, fail to notice that the response was not actually highlighted, and then move on to the next question. In order to avoid this problem, many online forms require subjects to answer every question before the form can be submitted. Yet ethical guidelines require researchers to allow subjects to skip questions they do not wish to answer. How can you strike a balance between allowing respondents to skip questions that they wish to skip, yet guarantee that subjects completely answer all the questions they wish to answer? Perhaps the best tactic is to require respondents to answer each question, but include a response option for each item that allows the respondent to skip the item. For example,

Withdrawing. On each page of your online questionnaire, you should include an option that will allow the respondent to withdraw from the survey. If the respondent chooses this option, all responses from that subject -- even data that you may already have captured and stored in your database -- should be discarded.

Confidentiality and Data Security

Some respondents may fear that their answers on your online survey could be traced back to them via their e-mail address, their IP address, or other information that you could attempt to capture from them while they are visiting your Web site. You should reassure respondents that you will not attempt to capture information that they do not voluntarily provide. In addition, if you are asking about sensitive information and you request respondents' e-mail addresses for any reason, you should specify that you will save the e-mail addresses in a separate file from the other responses, such that a respondent's specific responses cannot be linked back to their e-mail address. This is particularly important if you are gathering data on illegal activities, which may be subpoenaed by government officials; keeping respondents' e-mail addresses and other identifying information in a separate file, and destroying the file after it is no longer needed, is the only way to ensure respondents' confidentiality.

It is impossible to guarantee complete data security while respondents' answers are in transmission from their own computer to your server. In most studies, it is extremely unlikely that anyone would wish to "tap" the responses as they flow to your server. If privacy and security are of particular concern to your respondents, however, you may look into the possibility of encrypting your data flow. If you are gathering data on illegal activities, you must be particularly careful about your respondents' privacy; you should caution your respondents about the risks involved in participating in the study, and you should allow them to reply to the survey in a more anonymous manner. In his article on collecting online data from illicit drug dealers, Coomber (1997) suggests that respondents can either use an anonymous terminal (for example, a computer in a public library or cyber-cafe) where the response cannot be traced to an individual, or print the questionnaire and send it to the researcher via regular mail.

Please be aware that despite your best efforts to protect your respondents' identities, it is nearly impossible to fully guarantee your respondent's anonymity; thus, in most cases you may only guarantee confidentiality.



Information Technology Services. Infrastructure. Innovation. Integrity.

Last updated June 30, 2008.
Copyright © 2001-16, Information Technology Services at The University of Texas at Austin.
All rights reserved. For privacy concerns read our privacy policy.

To submit questions or comments regarding this page, use the online Comment Form.