What are panels


A certain constant group of respondents (persons, companies) who are questioned about the same thing over a longer period of time. Example: GfK household panel (society for consumer, market and sales research), which examines the purchases made by households in retail (retail).

As a special form of primary research, it represents a constant sample of selected, representative respondents who regularly provide information over a longer period of time about an object that is basically the same. The most important panel types are consumer panels (also called household panels or individual panels, comprise a sample of households or individuals), trade panels (wholesale and retail panels), company panels (sample of all / selected companies or individual sectors, e.g. cosmetic panels) and special panels (e.g. local test market panels) . Online panels enable surveys to be carried out quickly and inexpensively via the internet and also limit tests with internet users registered with the online panel offering market researchers. Methodological problems of the panels (limits of a panel) arise firstly with a large sample (panel size). In consumer goods and durable goods panels with 5,000 / 10,000 test subjects, problems with maintaining representativeness under the keyword of panel mortality (leaving the panel for reasons such as moving, lack of time, death) emerged. Failures of this kind are countered with a panel rotation (failed participants are replaced by new participants). Second, with the panel effects (changes in behavior are induced by repeated surveys, e.g. Ā»overreportingĀ«). A third problem is the degree of market coverage (coverage) of the panels, because panel results cannot, for example, represent all households or retailers, as the willingness to participate often does not correspond to a representative sample structure.

Repeatedly querying the same group of participants about the same object to record changes. Among other things, market share, buyer structures and brand loyalty can be read from the panels. Representation can be affected by panel mortality and high refusal rates. A change in purchasing behavior due to the observation and a decline in willingness to participate or the thoroughness of the entries result in the panel effect. Application examples are the determination of audience ratings and the recording of food purchases.

In market research, the panel is a constant group of respondents who basically remain unchanged for the duration of the survey. This constant group of respondents as the underlying survey procedure has led to the name panel procedure (English panel = list of names).

The term panel has been adopted from Anglo-American empirical social research. A panel is a specific, basically constant, representative group of respondents who are surveyed about the same subject over a longer period of time, either continuously or at specific time intervals. This definition makes clear the objective that panel research primarily pursues: the recording of movements or changes over time. The most important types of panels can be shown schematically as follows, depending on the task (see p. 336):
The most important panel variants are the household panel and the retail panel.
The aim of the retail panel is to determine the development of the sales situation with a structurally consistent representative sample of sales intermediaries (trading companies) for specified brands, products and product groups. The data acquisition is carried out by employees of a market research institute in the same way as an inventory process. Inventories, incoming goods and sales of goods are visited every two months

Certain, constant, representative group of respondents who are questioned continuously over a longer period of time or at certain intervals about the same subject in principle.

is a constant group of respondents who are questioned several times on the same topic at different times of the investigation. Panel surveys are not an independent survey technique, but are a special form of survey that can be carried out in writing, orally, by telephone or with the aid of a computer. See also market research methods and market research (with references).

Standard information service that is very important for market research. A panel is a specific, constant group of addressees who repeatedly (at regular time intervals) surveys on (in principle) the same subject of investigation. This can be done through oral, telephone and written questioning, but also through observation. The survey group (panel participants) depends on the subject of the investigation; In principle, any type of information group comes into question, such as private individuals, private households, self-employed traders, companies. The conceptual feature of the constant circle of respondents must not be understood in a too narrow sense. Panel failures or changes in the population (e.g. due to changed purchasing habits) must result in adjustments to the sample that are intended to represent this population. The subject of the survey, which is basically the same, also only applies with restrictions in practice. Due to product innovations, variations or eliminations, the object of investigation can change significantly over time. The special characteristic of panel research is that the same sources of information are used repeatedly, on the same topic. Panel surveys are not to be confused with wave-shaped surveys that contain the same topic repeatedly, but in which the respondents themselves are not identical. The panel research thus records movements or changes over time, namely with the same respondents. The faster the objects of investigation change over time, the more necessary a continuous observation of the changes that have occurred (strategic market research). This circumstance - typical for modern market events, but also for many socio-political developments - has inevitably led to a strong expansion of panel research in the last few decades. Due to the mostly considerable organizational and start-up costs as well as the high ongoing effort, panel surveys are mainly carried out by market research institutes, in Germany especially by GfK and Nielsen. Once panels have been successfully installed, they are often an instrument with continuous sales and income for the institute. Because customer participation in panel surveys usually takes place Usually long-term, corresponding degrees are concluded for at least one year. The more customers participate (exception: exclusive panel contracts), the more severe the degressive effects. Decisive for the information value of panel surveys are the degree of representation, the accuracy of the data collection and processing as well as the speed with which the data is available. The selected time intervals are primarily based on the dynamics of the investigation conditions, but also on the technical and cost aspects of the survey. Since the first application of the panel method to questions in the field of market and opinion research, a whole series of panel forms has emerged, of which the consumer panel and the retail panel are undoubtedly the most important; Other forms and variants relevant to questions of sales economy are shown in Fig. 1. Consumer panels are all those panels in which the group of respondents is made up of end consumers. So it includes either individuals or households. The individual panel will always be suitable when information is required that directly relates only to the individual (e.g. personal needs for cosmetics or tobacco products). In the case of the household panel, the focus is accordingly on the procurement of household-related data (e.g. food purchases). Another distinction relates to the product groups to be examined via the panel and differentiates accordingly between the consumer goods panel and the consumer goods panel. Since a consumer panel cannot contain the entire relevant population, a representative selection is made. The characteristics that are used for the formation of the sample and thus for the structure of the panel sample (sample) are determined by the requirements that the purpose of the panel has to meet. The same applies to the sample size. In practice, a number of participants between 2,500 and 10,000 has so far proven to be sufficient in Germany after weighing up cost considerations and significance claims. When recruiting respondents to participate in a panel, higher refusal rates are naturally to be expected than, for example, in the case of a one-off interview. Almost all institutes therefore offer rewards as an incentive, be it as ongoing remuneration for each report sent, as a one-off amount per year and / or as automatic participation in the raffle for high-quality items. Behind this is also the effort to choose a form of reward that does not lead to a deviation of the panel from the population from the outset. However, problems of representation arise primarily in other respects. For one thing, household panels do not cover all households in Germany. Your representation only applies to private resident households; Households from foreign nationals, who are often unable to take part in the panel due to poor language and writing skills, and institutional households are usually not taken into account. On the other hand, failures are replaced over time, but only the remaining part - the so-called continuous panel mass - is actually identical over a longer period of time. After one year the total panel mass is approx. 80%, after 2 years approx. 70% and after 3 years only approx. 60% of all panel participants. The collection of data at the consumer panel usually takes place in the form of periodic written reports by the panel participants. The institution submits the registration form or report documents as well as the other required funds (postage paid envelopes, etc.) before the reporting date or collectively for several periods. For example, the members of the G&I Household Panel receive a quarterly so-called panel calendar, from which they take and fill out a sheet every week. At the consumer panel, the central subject of the query is primarily the purchases of certain product groups listed in the panel. The respondent must indicate: purchased product, brand (manufacturer), date of purchase, type of pack, size, number, price (individual and total), place of purchase (type, name, trade group), place of purchase, special information. The socio-demographic data of the panel participants provide further important information. When recruiting and at the beginning of a new calendar year, the panel participants are asked to provide information on certain socio-demographic criteria; These include the most important characteristics of the household size, the number of children, the age of the housewife, the schooling of the head of the household, the affiliation to a certain social class, the income per capita and the total household net income. The inquiries about purchases are sometimes followed by special surveys, such as media behavior in the period in question, in order to create appropriate starting material for later special correlations based on the single-source principle. After the information from the returned report sheet has been transferred to EDP-compatible data carriers, checks are first made for formal and logical errors. If there are any, they may need to be eliminated by interviewing the relevant sender. Following this, the panel data are then checked for consistency and logic in terms of content using appropriate test programs. The actual evaluation can be divided into 2 areas, the standard panel evaluations and the special panel analyzes. Standard panel evaluations are used to continuously monitor market developments and include tabular preparation of the individual report data. In addition to pure total counts, this includes in particular cross-evaluations. Fig. 2 gives an overview of the possible breakdowns and combinations, whereby a two-dimensional evaluation is assumed in the diagram. As a rule, however, the tables are created in several dimensions (up to 5 dimensions), such as the purchase quantity according to the place of purchase according to location in areas. Special panel analyzes are evaluations that are not carried out as part of continuous reporting, but rather aperiodically on the occasion of specific market events or in the context of decision-making about new marketing ideas by the institute and are remunerated separately. In detail, the range of services of the special analyzes can be characterized by the following evaluation options or objects: shopping intensity,> brand loyalty, cumulative buyer or repurchase rate (new product sales forecast), demand coverage, buyer migration, (gain-and-loss analyzes), introductory analyzes, Promotion analyzes, combination analyzes, correlations, price class analyzes. From a strictly methodological point of view, the group of panel participants would have to remain constant over the entire period of the survey (consistent panel mass) and of course present an exact image of the survey as a whole (representation). In practice, however, both requirements do not arise automatically; Corrections are therefore constantly required to reduce the errors caused by this. A group of 4,000 to 10,000 households is inevitably exposed to a certain panel mortality, i.e. a continuous fluctuation due to birth, death, marriage, relocation, etc., quite apart from the possible refusal of further participation. Each panel is therefore run with a certain reserve, i.e. H. In addition, a group of people is questioned in exactly the same way, from which any gaps in the panel mass relevant to the evaluation can then be filled. In practice, possible distortions of the sample results due to panel mortality are countered by working with a mixed random and quota procedure when selecting the panel sample. This means that the panel participants selected at random will be replaced in the event of their departure by participants from the reserve pool who have the same relevant characteristic structure with regard to their purchasing and usage behavior as the withdrawn participants. Basically, the panel becomes a quota model over time, even if the available reserve panel households are selected at random. Further causes for the continuous refreshing of the panel (panel rotation or artificial panel mortality) are occurring learning and awareness processes (panel effect) as well as the overreporting and underreporting of the panel participants. Overall, this makes it clear that a panel continuously requires extensive control and support. It ranges from checking the individual report information for completeness, meeting deadlines and formal correctness to ongoing monitoring of the feature structure of the panel and accordingly includes balanced contact with the participants, sufficient motivation and other "panel maintenance" measures. The particular importance of the consumer panel as a survey instrument in the context of primary research does not need special emphasis. Its data represent an almost indispensable basis for marketing planning and marketing controlling, especially for the branded goods industry. The fact that the methodological problems of the instrument could be adequately overcome is confirmed not least by the fact that panel research has shown continuous growth rates in the past few years. The previous form of the written entry and the postal return of the report sheet is of course not the optimum in terms of accuracy and speed. As a result, considerations and attempts are constantly being made as to how these processes could be improved through the use of modern technical means of communication. The Btx system comes into question as a data acquisition instrument. Direct data entry through such a tool would significantly increase the chance for faster reporting. So it will essentially depend on how quickly Btx asserts itself in private households so that the demand for representation is met.On the other hand, the increasing spread of private personal computers could enable them to be used as data acquisition devices. Another, technically much simpler alternative would be the telephone announcement or inquiry. This hardly results in any time savings, since daily communication cannot of course be expected. In addition, the risk of transmission errors and the costs increase. Technically attractive solutions for increasing the accuracy and speed of reporting in the context of consumer panel research are home scanner panels and regional scanning.

Literature: Hansen, J., Das Panel. For the analysis of changes in behavior and attitudes, Opladen 1982. Sedlmeyer, K.-J., Panel information and marketing decision, Munich 1983.

Previous technical term: pallet rack | Next technical term: panel system research

Report this article to the editors as incorrect & mark it for editing