This week we continue to look at Using qualitative and quantitative research in your quality management system.
This technique is a one-to-one discussion between a skilled interviewer and a single respondent. It is difficult to be prescriptive here, as it is quite possible for interviews to investigate a narrow topic in depth, or a broad topic in less depth.
Quite often an interview will be timed to last for about an hour, but it could go on longer if the respondent is interested and willing. In-depth interviews are useful when probing motivations, attitudes and needs.
They are advantageous when the subject matter is sensitive, confidential or embarrassing, when social norms may be in question, or when a detailed understanding of a person's behaviour is needed. They are most often used with professional consumers, such as industrial buyers.
The method is particularly valuable when it is important to get at subtle, idiosyncratic differences and shades of meaning. It has been found to generate more and higher-quality ideas than many other methods. In-depth interviews are more expensive on a per person basis than focus groups.
The interviewer must be highly skilled, and even then may experience difficulty if more than four or five hours a day are spent interviewing. Another problem is that the large amount of disparate data generated can lead to difficulties in analysis.
This is because the researcher has to find a way to make sense of a number of interviews which may have covered slightly different areas, in different ways, depending on what the respondent wanted to say.
To overcome this problem, some interviews will use a checklist of questions to ensure that they cover basically the same areas each time, although this can lose some of the depth and original ideas and thoughts which respondents might have. As with many aspects of research, a balance has to be struck.
The premise which underlies observational techniques is that the best way to find out what people do is to watch them doing it in natural settings. Observation does not rely on self-report, or on co-operation from consumers. It may also be the only way to gather information about behaviour in situations in which consumers are not themselves aware of their own behaviour.
Observation can be either with or without the knowledge (and consent) of the observed, and may be direct or indirect. For example, you could watch how people go about making a purchase, or look at the results of their having made the purchase. You could record only certain predefined aspects, or everything that happens.
Underhill (1999) uses the technique of observation extensively in the USA to help retailers make decisions about where to stock merchandise and how to design their stores. For example, a major jeans manufacturer wanted to know how its product was sold in department stores, so in one weekend Underhill's company observed what was happening in four stores. It tracked a total of 815 shoppers and observed many more, using both video and time-lapse cameras.
By the time the study was complete, it had analysed the percentages of customers who used the various paths into the jeans section of the store. Once that had been worked out, it was clear that many of the signs were badly positioned.
Some customers were interviewed to reveal how their attitudes and opinions correlated with their behaviour, for example to see if young shoppers with good education levels, who said they depended on brand names when choosing jeans, read the price tags as well.
The study was much broader than described here, but you can see the power of observation from this example. You can also see that this observation study generated both qualitative and quantitative data, which were analysed together to gain even greater insight into what was happening.
Observation on its own cannot reveal the feelings, attitudes and beliefs that underlie consumers' behaviour. Further, many situations are not easy to observe in a systematic way — it would be unwise to try to observe the use of toothpaste in a truly natural setting!
Both the primary and the secondary data we have discussed previously have to b be generated in the first place. Research methods can generally be divided into qualitative methods and quantitative methods.
In simple terms, qualitative market research methods are used to explore the attitudes and behaviours of customers; quantitative methods apply numerical values to measure aspects of customer behaviours and attitudes.
Qualitative methods are of most use when you want to discover those attitudes, beliefs and behaviours which people would have difficulty expressing in precise terms. Qualitative data is generally more impressionistic than quantitative data, and so its usefulness very much depends on the skill of the researcher in gathering, analysing and reporting it.
If the hotel garden conversation we considered earlier had been recorded, it would have been a good example of the collection of qualitative data. Quantitative methods, on the other hand, begin with the premise that aspects of human behaviour, attitudes and the like can be measured. It is also assumed that the numerical values associated with a respondent's answers can be analysed using statistical tools.
Some quantitative methods assume that people can, and will, respond honestly and accurately to questions in the form of a questionnaire, an interview or a test. Quantitative techniques may be very expensive if a large number of people have to participate in order to provide sufficient statistical significance.
If the hotel in our example provided a short questionnaire to a sample of its customers over a specified period, in which customers were required to rate their satisfaction with various aspects of the hotel's service, that would be quantitative research.
Some researchers like to think of qualitative studies as a preparation for quantitative studies. In this case, the results of the qualitative research are used to help identify exactly what should be measured by the quantitative research. However, both types of research can be useful in their own right. In general, qualitative techniques involve the gathering and analysis of non-numerical data.
Focus groups are especially useful in developing new, or modifying existing, services or products. They are used in most industries and sectors. Small groups of individuals (5 to 12 paid members) are chosen to represent the characteristics of customers In a particular market segment. Their discussion is led by a moderator, often a psychologist or trained specialist, and is usually recorded on video or audio tape.
Its main points are drawn out by the moderator from a transcript of the discussion, and this serves as the basis for the report to the managers who make the subsequent decisions.
Focus groups can be constituted quickly, and they are normally relatively inexpensive. Multiple focus groups, each representing a single important market segment, cost more.
Group members can give deep, rich, subtle and well-developed opinions about the topic of the research, enhanced by the debate which occurs between them. Another advantage of focus groups is that the decision maker can either watch the group through a one-way mirror as it meets, or replay the group session later on tape. This helps the decision maker to stay in touch with consumers in a way which might not otherwise be possible.
Focus groups can be run in a more informal manner to help get a feel for an issue or a problem. In this case, a good facilitator and a notebook may be all that is required.
The major disadvantages of focus groups are that highly skilled, and therefore expensive, moderators may be required to run them; good sampling is difficult to arrange; moderators may introduce bias into the process; and group dynamics are very likely to enter into the process, thus leading the group to express views that the individuals would not express on their own. As with all qualitative techniques, focus groups do not produce statistically significant results.
Next week we will look at the role of in-depth interviews in qualitative research methods...
Last week we looked at the role of external data in quality management systems. This week we look at the role of primary data.
Gathering primary data can be complex, often requiring special skills or knowledge. Many organisations contract specialist research agencies to carry it out for them. However, it is often important for front-line managers to be consulted about the questions to be asked and about the use to which the information obtained is to be put.
The aim of this section, therefore, is to help you to appraise the market research produced by others. We look first at three main sources of primary data before turning to the detail of the quantitative and qualitative methods of generating it.
Syndicated research Some market research companies have ongoing research programmes, the results of which they sell to a number of clients. Some of them can be standard research, such as the A.C. Nielsen store audits, which provide information on retail purchases by consumers in a number of countries, or the TGI (Target Group Index) of the BMRB (British Market Research Bureau), which has followed the fortunes of some 5,000 brands in the UK for more than 20 years.
Syndicated research has the advantage of shared cost, but its value depends on the standards of quality and rigour various companies have. Some syndicated research can be ad hoc: a research company, perhaps specialising in an industrial field, sees a topic which it believes will be of interest, conducts the research and sells the results ‘off the shelf’.
Some research organisations with ongoing programmes, especially those conducting opinion polls (for instance Gallup) will sell ‘space’ — or more accurately interviewer time — on the back of their surveys. Clients can ask for one or two simple questions to be added at the end of the main 'omnibus' survey of a large sample.
One approach to researching customers is to track the purchasing behaviour of a carefully selected panel of individuals. The data from these individuals' behaviour — as opposed to their opinions — reveals important information about repeat purchasing and brand switching, which is almost impossible to obtain with other methods.
Panels are used to generate television viewing numbers. These numbers are important to the television channels, but perhaps more importantly to the advertisers who pay for space on these channels. Households are selected so that a panel adequately represents the entire population.
For example, there should be the same percentage of single households and households with one child, two children and so on in the panel as there is in the population as a whole.
Each panel household is given a device to record its members' presence in the room when the television is on. Each member of the household has a button to press, or a number to key in to register that they are in the room. In many cases the channel being watched is automatically recorded.
This system claims to record what people watch, rather than what they say they watch, although it does not record their concentration on the programme (or more importantly the advertisements) while they are in the room.
This is the staple diet of the market research industry. A specialist custom research company is commissioned by a client to undertake a specific piece of research. The research company then accepts responsibility for all aspects of the research. There are companies which cover, or claim to cover, all types of market research. On the other hand, many companies specialise in particular fields.
For example, there is usually a distinction between those which specialise in consumer and in industrial fields. Equally, there are clear distinctions between those involved in retail audits and those conducting questionnaire surveys on individual consumers, and between those running focus groups and those carrying out in-depth psychological interviews. These will be discussed in a later section.
‘Desk research’ is based on data gathered elsewhere and made available. Sometimes it may generate all the information that is required, particularly in the initial stages of collecting information.
There are six general sources of desk research. These are:
Computerised databases. A computer database can be a source of information on a wide variety of topics. The problem is not the scarcity of information but the amount and reliability of it. Finding exactly what you want from the vast amount of data available takes time and skill in working with search engines.
Associations. Almost every business, service or charity belongs to some sort of trade association which, as part of its normal activities, keeps records which may be of use in market research. For example, the Charity Commission with which UK charities have to be registered provides consolidated information about donations. Information from associations is often available to non-members for a modest charge. Trade unions may also provide useful secondary data.
Government agencies. Government agencies — local, national and international — produce demographic data, sales data, employment data, import and export data, and special reports on industries and markets. For example, the Office for National Statistics website contains a useful index of statistical bulletins to help you find topics of interest, and other governments offer similar indexes, as do some international agencies.
Syndicated services. Information on almost every conceivable topic is produced for sale by market research services. Again, a wide variety of sources are available, but a good starting place in the UK is the Market Research Society’s Knowledge Centre, which is available online. Once more, the equivalent societies in other countries are a good starting-point for your search.
Libraries. Librarians are trained in the retrieval of specific information and can help you to find what you need, on paper or electronically. Many libraries offer inter-library loan services.
The media. The general news media, daily newspapers, television and trade press are rich sources of information. Reading is an effective way of keeping abreast of developments in business and management in general, and in your sector in particular. There are media covering most industries, professions and sectors from advertising and aromatherapy through to zip-fasteners and zoology, all of which can yield useful information to managers working in relevant organisations.
Secondary data is typically used to enrich the context in which market research takes place. You need to know who produced the data, why, how and when. This knowledge will help you to estimate the relevance of the data for your decision.
For example, if you want to rely on government information to help you analyse foreign markets, you need to be aware that the statistics about different countries may have been gathered with a different purpose in mind.
Further, they may not have been collected using methodologies that are consistent from country to country, nor may they all have been collected at the same time. All these factors could have an effect on the validity and comparability of the data.
An example of this is the way data on inflation is analysed in different countries. In the UK the inflation rate is calculated excluding the price of cigarettes; in Ireland the rate includes the price of cigarettes. Thus, an increase in government duty at budget time in Ireland can radically alter the headline rate of inflation.
Next week we'll look at the role of Primary Data...
Here you'll find the latest blog articles on all things compliance, particularly focussed on quality, environment, health & safety, rail, and information security.