An online panel is a group of registered people who have agreed to repeatedly participate in online studies. The panel data collected via the Internet are mainly available for market research and social science data collections .
The process begins with the registration on the respective website of the online panel. This automatically gives permission to the organization that will later collect data on the Internet to be repeatedly contacted for market research purposes . The organizations select the panel participants in such a way that the sample obtained is representative of the population to be examined . In addition, the selection can be made either randomly from all participants who are already registered or based on quotas previously set in the system, if this is advantageous for the study to be carried out. For example, only women of childbearing age can be asked about a pregnancy preparation.
In order to motivate the relevant people to participate in an online panel, incentives such as money, loyalty gifts or vouchers are often created. However, these should be designed in such a way that they do not influence the behavior of the participants. Too small an incentive (or no incentive at all) could prevent people from participating in the panel at all, while too great an incentive could distract people from the survey and thus distort the online study.
If the respective participant was then successfully registered, he is saved in a database from which the samples for the individual surveys are then drawn. He thus becomes part of a more or less large pool of people who voluntarily make themselves available to such online surveys.
Compared to conventional panels
The classic panel examination is understood to be the multiple collection of the same variables on the same examination objects at different times. This means that a group of people is asked questions on the same topic at regular intervals.
The form of survey of the online panel arose from the growing medium of the Internet and the associated expansion of market research into the Internet.
The term online panel is used to designate classic panels operated via the Internet as well as online access panels. The former are structured like classic panels, with the difference that the data is collected here via the Internet. With an online access panel , the understanding of the classic panel is expanded. An online access panel means the multiple collection of different variables on the same examination objects at different times. In addition to the previous aspect of the constant survey sample, there is also the fact that questions are asked on different topics. Due to this constant variation of topics, the online access panel for market research can be used as a versatile pool of people.
Advantages of online panels
The advantages of online panels are primarily based on the general advantages of the Internet. The following advantages are particularly relevant:
- Cost minimization: In contrast to conventional classic panels, online panels are much cheaper to implement, since all the procedures such as surveys, data collection and customer data can be carried out online. However, due to their structure and especially their maintenance (see disadvantages of online panels), they are often more expensive than other online survey methods.
- Short implementation times: Studies that are carried out using online panels can usually be implemented very quickly, which is mainly due to the fact that the participating people are already known and therefore easy to reach.
- High flexibility: The inclusion of the Internet in the research to be carried out results in increased flexibility both with regard to the examination location and the examination time.
Disadvantages of online panels
- Panel effect: the behavior of the people participating in an online panel can change significantly over time because they are aware of their observation . The change in behavior can happen consciously or unconsciously. Panel effects are a major problem because they impair the meaningfulness of the results. The easiest way to prevent this is to regularly recruit new participants.
- Establishment and maintenance: Before active research can even begin, a system must be created that meets the requirements. What is meant here are, for example, online databases for survey data and customer data. In addition, participants must be recruited using complex criteria. After all, the online panel must always be kept up-to-date and representative, which is often associated with considerable expenditure of time and money.
- Representativity: The biggest problem with online panels is the lack of representativeness. This circumstance is mainly due to self-selection . Due to the conscious decision of the participants to take part in the panel, the sample is not representative of the population. Usually, intensive internet users are clearly overrepresented. In practice, therefore, one often works with quotations when selecting the participants or weightings when evaluating the data in order to achieve representative results at least for certain quota characteristics.
The representativeness is particularly important with regard to online panels. An online panel is generally considered representative if the results of the sample can be extrapolated to the population . This is the case if the results of the panel participants are not only valid for the respective group surveyed, but can also be assumed to be generalizable. If a sample is not representative, the result is distorted and an extrapolation of the research results to the population is no longer possible or no longer useful.
Such a distortion of the results, which leads to the sample no longer being regarded as representative, can arise either due to the non-participation of the investigators or due to selection processes. The non-participation of the investigators, such as inaccessibility, refusal to participate and the like, only play a subordinate role in the online panel, as all panel participants take part consciously and on a voluntary basis. The selection processes, meaning here the criteria according to which the investigators are selected, have a decisive influence on the representativeness. They can be divided into the two sub-areas of self-selection and the population of online panels.
Self-selection means that the initiative to participate in an online panel or for other research purposes does not come from the social or market research institute, but from the respondent himself. So this one is not selected, he chooses himself, so to speak. This conscious decision to take part in a certain study leads to a distortion in the form of an overrepresentation of certain groups of people and the sample is no longer considered representative. Self-selection also occurs more frequently with controversial topics than with common topics. For example, a disproportionately large number of strict opponents of nuclear power may take part in the survey on the construction of a new nuclear power plant .
The population of online panels
The cause of this distortion trigger is the internet itself, since the population of online panels consists only of internet users. Since the Internet is absolutely necessary to take part in an online panel, a significant part of the total population is excluded, namely all those without Internet access. Even if the proportion of people without internet access is decreasing, it still represents an important part. Studies have also shown that among all internet users there are proportionally more younger, higher-income, better-educated and more men than women compared to the general population . Due to the steadily advancing penetration of the Internet, the representativity between Internet users and the general population will converge more and more in the next few years, which will weaken this distortion.
To ensure that an online panel is representative, it must be ensured that the results are not distorted. In order to keep these distortions as low as possible, it is necessary to plan and document the entire recruiting process . Because especially with online surveys, people with a strong internet affinity are more motivated to take part than those without internet connection. As a rule, it is advisable to recruit the relevant test subjects not only online but also offline. In general, it can be said that people who are willing to provide information should be sought who can be reached and addressed in the further course of the investigation on the Internet.
Active and passive recruitment
- Active recruitment: With active recruitment , the participants are contacted directly by the respective operator of the panel by email or by phone and are motivated to take part in an online panel. This procedure can be used to determine exactly which people should take part in the online panel. In this way, a specific, self-contained group of participants can be set up.
- Passive recruitment: With passive recruitment, the participants do not get their information directly from the operator of the panel, but from external sources such as search engines , links to websites, articles in magazines or from friends and acquaintances. By doing this, it is not possible to predict exactly which people will ultimately sign up.
Online and offline recruitment
Online recruitment: Online recruitment is understood to mean the recruitment of participants using common online media such as e-mails, banner advertising or newsletters . The four most common channels are:
- Internet users register directly on the website of the respective panel operator
- Internet users are randomly selected from a customer profile by Internet providers or Internet service providers and addressed
- Visitors to selected websites are addressed using pop-up windows
- Advertising banners on selected websites refer to the corresponding panels
- Offline recruitment: Offline recruitment is understood to mean the recruitment of participants using conventional, non-internet-related means, such as by post or by telephone. Face-to-face recruitment is also offline recruitment. This method can be advantageous if you want to avoid people with a strong Internet affinity for certain survey topics.
The online panel should always be kept up-to-date and representative. Furthermore, it should be prevented that there is a change in the behavior of the participants and thus a panel effect. It is therefore important that new participants are constantly recruited so that there is a sufficiently large pool of people available at all times .
ISO standard for access panels
In 2009, the international standards organization ISO published the ISO 26362: 2009 standard Access panels in market, opinion and social research - terms and service requirements ( English Access panels in market, opinion and social research - Vocabulary and service requirements ). This standard is applicable to all types of access panels, especially those that are recruited and used online. In addition to this specification of the scope and terms and definitions, the standard contains general requirements and specifications for
- Organization and responsibility
- Recruiting new panelists
- Confidentiality and Transparency
- Methods of recruiting
- Sources of Recruitment
- Validation of identity
- Structure and size
- Handling of the profile data of the panel participants
- Management of the access panel
- Maintenance and updating of the profile data of panel participants
- Pretest and translation
- Invitations to participate in research projects
- Validation of the data
- Reporting to the client
- Duties towards the panelists
- Obligations to the client.
The ISO standard affects all work and process flows in research projects via access panels, from the entire responsibility and management structures to the recruitment of panel members and the validation of the data sets to reporting.
Based on ISO 26362, interested access panel providers have been offered certification since 2010, during which the quality criteria of the standard are checked. Austrian Standards plus GmbH is the organizational sponsor of the certification program. The certification program was created by the Working Group of German Market and Social Research Institutes (ADM) , the German Society for Online Research (DGOF) and the Association of Austrian Market Researchers ( VMÖ ) together with Austrian Standards plus Certification. So far, several institutes in Germany and Austria, including MindTake, Marketagent, Norstat Germany (formerly ODC Services), respondi, GfK Austria, have subjected themselves to the conformity assessment according to this program.
- Göritz, AS & Moser, K. (2000). Representativeness in the online panel. The Market, 155, 156–162.
- Petrovic, O. (2009). Determinants of an optimal online panel design. GRIN publishing house. ISBN 3-640-33857-X
- Theobald, A., Dreyer, M. & Starsetzki, T. (2003). Online market research: Theoretical principles and practical experience (pp. 228–238). Gabler Publishing House. ISBN 3-409-21781-9
- Technical article on the topic of: Research & Results