Online Research

The use of the internet for quantitative research emerged and grew in the 1990s. In the early days there were doubts about its efficacy — Was it representative? Yet, as confidence in its use grew, online research experienced explosive growth.

Online — Advantages

Today, according to the ESOMAR industry reports, online is the most widely used method worldwide for quantitative research. Its benefits, many of which are listed below, give it a distinct advantage over other methods.

• Online research improves respondents’ level of interaction and engagement through use of multimedia content — image, video, sound, animation and text. User friendly, engaging features include elements such as sliding scales with smiley faces, 3-dimensional images and interactive exercises. For instance, in some concept clinics or pack tests, respondents are able to see how the choices they make change the visual image of the product or packaging. The richness and depth of participants’ responses improves through interactions of this nature that trigger higher levels of involvement.
• The depth and quality of responses also improves as respondents have more time to think and react to open-ended questions. Compared to face-to-face or telephone interviews, in an online setting, respondents feel less pressure to respond quickly or in a manner that is politically, socially or intellectually correct.
• Online is better suited for obtaining sensitive information since respondents may remain anonymous.
• Being computer-assisted, online research supports better questionnaire controls. It employs software to control the sequencing logic and flow of the questionnaire, check sample quotas, and so on. 
• Faster turnaround is achieved as online research is not constrained by the limitation of a physical field force. Respondents can respond concurrently in real time. And as they click their submit buttons, the data is acquired and, in some instances, auto-processed. The resulting improvement in turnaround is a key advantage considering the pace of developments in markets. (The speed-to-market especially for technology products and consumer durables has shrunk from years to months. Marketers of these products, can no longer afford delays in execution due to time consuming market research programmes.)
• Automation and consistency in delivery eliminates interviewing bias in online surveys. (Though this is not likely to be as big an issue since we use more close-ended questions in quant, interviewers can sometimes subtly bias research through their personality, appearance, body language or tone of voice.)
• Respondents’ interaction with the specific question pages and stimuli can be tracked in terms of time spent and click-through rates. This may be used for optimization of questionnaire designs. It reveals the more interesting or engaging aspects of the questionnaire. It also helps identify “professional survey takers”.
• Online research is not constrained by geographical boundaries or barriers. Whereas certain residential areas may be inaccessible for DTD interviews, this is not an issue with online.
• The participation of respondents such as specialists, professionals or new mothers who would otherwise be hard to get, can be facilitated through specialized access panels.
• Cost savings arise from multiple areas — no fieldwork, savings in data entry and coding. Where applicable there are also savings on the construction of physical stimuli or prototypes, which are substituted by visually appealing graphic images.
• The reduction in costs also affords greater flexibility. It is feasible and affordable to test different variations of questionnaires, stimuli and concepts. It is possible too to change or add questions on the fly.
• In terms of logistics and costs, multi-country studies benefit greatly from the borderless nature of the internet.  They are easier to coordinate and execute simultaneously across countries. Online also eliminates the need for the physical presence of teams in each country.

Online — Limitations

For the above reasons, online research is the preferred method that market researchers adopt, when feasible. There are however limitations; and the one that is top of mind is representativeness. In quantitative research theory, the notion that samples should be random and representative has prevailed for a long time. Considering that online samples are neither random nor representative, raises a number of questions both from a theoretical as well as a practical standpoint.  For instance, it is also debateable whether conventional methods of data collection are random or representative.

The response rates for conventional offline surveys are low and are getting lower as people’s hectic lifestyles undermine their willingness to answer questionnaires on doorsteps or over the telephone. DTD surveys also exclude people who dwell in restricted areas. Moreover, surveys often need to be compensated by boosters or quotas. So, in reality there are no purely representative samples. Considering that some compromise must be made, decisions ought to be based on a clear understanding of what is required to meet the study objectives.

The vast majority of research studies tend to be broad-based and relative in nature. For instance — Which products do people like more? What attributes are associated with my brand? What are the factors driving customer satisfaction? The biases inherent in online are unlikely to affect the outcome of these studies, provided the research topic has no direct bearing on the internet.

Online samples tend to over-represent high internet usage segments (young, upper/middle class) at the expense of others. Research firms attempt to compensate for this to some extent by recruiting panellists in their access panels that better represent the population as a whole.

Besides representativeness, there are few more limitations of online that need to be taken into consideration:

• Like other self-administered questionnaires, online depends upon respondents being able to comprehend, on their own, what is required. Similarly, if the respondent’s answer is ambiguous, there is no interviewer present to seek clarification. It is important therefore that questions are clear and to the point. If the topic of discussion is complex, in-person methods may work better.
• The internet has led to the rise of “professional survey takers” — respondents who actively seek online surveys offering paid incentives and go quickly through the survey without devoting adequate attention to the questions that they answer. Web analytics makes it relatively easy for market research companies to spot these respondents. Most of them have procedures in place to identify the “professionals” and remove their responses from the study.

In summary, online methodology works best for research studies where the target respondents are available online and representing the entire population (online and offline) is not a necessity from the standpoint of data interpretation.  It is important that the questions are clear, concise and easy to answer. The over-representation of internet users will affect the results, particularly for topics that have some bearing on the internet.

Online — Methods

The most common online approach with established research houses is access panel based research. These are fully managed online panels comprising pre-screened individuals who have agreed to participate in research studies. Participants are usually incentivized by means of reward point schemes, sweepstakes or cash.

Online research may also be executed via a banner or a link that takes respondents to the questionnaire. While this approach is easy and inexpensive to implement, it offers limited controls on the sample, and representativeness is therefore a bigger issue. It should however work well where target respondents are visitors to a particular site.

Whereas banner-based research is open to all, a client list based research method is limited by invitation only to a targeted list of respondents. It is appropriate where the research requires purposive sampling.

Interviewer Administered Methods — Advantages

In-person methods are expensive compared to self-completion, yet they do offer a number of advantages, some of which are listed here:

• In-person interviews are recommended for complicated questionnaires because the interviewer can help with clarifications. Interviewers may also probe respondents in order to seek clarifications, or to overcome unwillingness to answer specific questions. Probing also helps enrich responses to open-ended questions.
• Skilled interviewers are capable of building interpersonal rapport thereby sustaining the respondent’s interest.
• In-person methods are required in surveying illiterate respondents.
• Interviewer administered questionnaires are also conducive for generating spontaneous “top-of-mind” responses and would therefore be recommended where this is needed.

In general, face-to-face interviewing has some advantages over telephone interviewing, but it is also more expensive. It is recommended for long and complex questionnaires and is required for research where the consumer is required to respond to sensorial stimuli. In addition, a controlled setting, as in CLTs, allows for a better managed environment, which is beneficial for some studies.