There are many reasons why the data owner may not want
to divulge further details or may not be able to help you.
One reason is that storage space costs money. They may be located in buildings that are secure and not easily accessible. They may only be available on microfiche, in which case, specific apparatus must be available for viewing and data transfer. Even software and hardware have both changed over the years. On the Internet, we may find that web addresses have been changed, password protected, or entire archives have been removed. A solution to some of these problems is to use the Internet Archive’s ‘way back machine’. The Internet Archive is building a digital library of Internet sites and other cultural artefacts held in digital form. Like a paper library, the archive provides free access to researchers, historians, scholars, and the general public (http://www. archive.org/). Very often web addresses that are not accessible, or have been removed, can be found by typing the address into the archive search form. The UK Web Archive has been collecting UK websites since 2004 in the same way (see http://www.webarchive.org.uk).
Things that must be considered in ‘reverse researching’
Item
Missing data Nothing is complete, but detecting what is missing can be informative
Mistakes Simple spelling mistakes may be indicative of a lack of attention to
detail; continue checking to identify any numerical errors
Poor expression
Poor expression is a problem, especially with texts that have been produced in a language in which the original researcher is not fully
fl uent. The sense may have become corrupted
Deliberate bias
The initial reason for any document is to achieve some objective. This tone can sway the desk researcher
Scales May bias the data in a specifi c direction
Table
3.11
© istockphoto.com/Stuart Miles
Part 2 Data collection
96
a term, enables policymakers to determine the most and the least popular decisions. Long-
term studies indeed show trends, but peaks and troughs can sometimes be attributed to
changes in the data collection method rather than real changes. For example, the 1881 and
2001 censuses cannot be compared without accounting for basic diff erences. The report-
writer must therefore be prepared to delve into studies far more deeply than might be
anticipated.
To assess the reliability of information, it is useful to look at the source and the context.
Table 3.13 shows a way to assess a particular piece of information. A score is given for these
two characteristics. If the total is six or above, then the information is likely to be trustworthy; if
it is fi ve or below, then use of the data should be made with caution.
Analysis for the purposes of the current research problem inevitably means that the
information may need to be reworked or combined in some way (see Figure 3.7). This
Raw and processed data distinguished
Examples of raw data Examples of processed data
Number of visitors to a store Tables
Customer comments List of comments classifi ed
Sales transaction reports Annual company report
Transcripts of conversations Focus group reports
CCTV recordings of sales encounter Quality assessments
Websites in other languages Translations
Table
3.12
Source and context appraisal
Appraisal of source (with score) Appraisal of context
Completely reliable (5) Confi rmed elsewhere (5)
Usually reliable (4) Probably true (4)
Fairly reliable (3) Possibly true (3)
Not usually reliable (2) Doubtful (2)
Unreliable (1) Improbable (1)
Reliability cannot be judged (0) Truth cannot be determined (0)
Table
3.13
Chapter 3 Secondary data
97
reworking or combining information from diff erent sources is a form of ‘data fusion’: a technique that has become more sophisticated and largely automated in recent years.
For any sector, the following information should be compiled to give a complete picture:
• Six top companies
• Six important trade press titles
• Three trade associations
• Two trade directories.
A web resource created for the UK market research industry http://www.mrweb.com/ sectors/ illustrates this well by giving instant links to sectors such as fi nance, pharmaceuticals, and many more (see Figure 3.7).
There is a distinction between synthesis and analysis that is worth stressing. ‘Synthesis’
is essentially description; there is little critical input apart from the decision about what to include and what to exclude. ‘Analysis’ implies creating an analytical framework that gives some structure that can be transferred to other evidence, to other cases. It implies the creation of a problem-solving tool. Many such tools have been adopted in the past: any matrix, a SWOT report, a PEST analysis, even the four Ps of the marketing mix.
Recording and reporting sources
In the business environment, reports are not expected to carry references to every statement made. This is primarily to remove irrelevant reading matter from reports, on the assumption that what is written is more important than who wrote it originally. This can
be quite alien to the academic who is well versed in the Harvard system of referencing, whereby any new fragment of knowledge is attributed to the originator. In many ways, business reporting is more diffi cult for the researcher because any point may be challenged
by users of the information; they will evaluate the usefulness of decision-making material in many ways, so you may well need to show the exact origin and method used to acquire the data. A second purpose for keeping detailed records is to make newer searches in a focused way and to avoid repeating abortive searches. For these reasons, it is crucial to keep good clear records.
There are two important exceptions to the attribution of sources: tables will carry a source with date and the report appendix should carry a summary of the main sources (and methods used). The ‘private’ records kept by the researcher should allow further detail to be provided when requested, or to extend a project beyond the initial phase. These records will be in the form of notes, photocopies, fi ling index cards; online searches will result in downloaded fi les that should also be kept in some form of system. The sources and dates of access should be recorded in all cases.
Part 2 Data collection
98
Explain the nature of secondary data
Secondary data are best remembered as ‘second-hand’, because such data are ‘old’ primary data.
This involves more than consulting documents; it also involves expert interviews with people.
Internal secondary data are generated by the organisation in question, whereas external secondary
data are created by agencies outside the firm. The term ‘tertiary data’ refers to indexes, citations to
articles, and summaries or abstracts; in simple terms, these are ‘third-hand’ data.
Outline the different sources of secondary data
Internal secondary data include: sales force records, transactions, competitor activity, customer
records, inventory records, cost of transport, previous research, and website visitor records.
External secondary data include government output, academic knowledge, and company reports.
Professional associations publish industry overviews. Market research report publishers have
1
2
Chapter summary
Purpose Why was it done?
Who paid for it?
What was the problem to be solved?
Population Who are the players in the sector?
Who was interviewed? How many?
What was the source of names?
Procedure What data capture method was used?
What quality control was there?
Does the method make sense? What is wrong?
Is it valid and reliable?
Publication Why was it distributed?
Who are the likely readers?
What decisions may have been made on the basis of this project?
Figure 3.7 Questions to ask about prior research
Chapter 3 Secondary data
99
reports on different market sectors. Geodemographic companies produce data from the census combined with other databases. There are also ‘consumer-generated media’ (CGM).
Show the benefits and the limitations of using secondary data Secondary data have various uses—they may answer the research question; they may also help to refine objectives, to design primary research, to assist in sampling, and to supply pre-codes for questionnaires. Secondary sources can be used in conjunction with primary research, for example,
to enhance understanding of and to confirm results. Secondary data are likely to be less expensive than collecting information freshly. They are relatively easy and quick to obtain: a unique benefit for international studies. However, secondary data have already been collected for something other than the current research problem. Therefore, they may not address the topic in question or may only provide part of the information expected. The information may not be accurate or it may
be outdated. Additionally, data may not actually exist, or may have restricted circulation.
List the features of a plan to carry out effective searches Three approaches are likely: visits to general and specialist libraries, contacting human experts, and carrying out computer searches. For all of these, the desk researcher must set limits on time spent, cost expended, number of sources searched, language used, age of sources, format of data, and methodology used. Continuous needs for data will make it feasible for the researcher to implement permanent information requests; conversely, the occasional requirement points towards gleaning information quickly. Desk research will start with the Internet, so the user must be familiar with search tools and techniques. To assess the reliability of information found,
it is useful to look at the source and its context. Use the MR mix to ascertain the original purpose, the population used, the procedure followed, and the reason for original publication. For any
sector, look for the top companies, important trade press titles, trade associations, and trade directories.
Review questions
1 You interview a journalist about reports he produced two years ago on the railway sector.
Is this primary or secondary data collection?
2 List the differences between internal and external secondary data. Explain whether one is better than the other and why.
3 Assess the advantages and disadvantages of using secondary data. What are the alternatives?
4 How can professional organisations be useful sources of data?
5 What are the main features of a desk research plan?
Discussion questions
1 Use the Internet to describe the market for baked beans in a particular country. You should provide market shares and brief descriptions of the major players.
3
4
Part 2 Data collection
100
2 Visit http://www.google.com You shouldn’t leave the Google pages. Look at the user
guide. Discover the services available and the user guides. Be sure to see Google Scholar
and Google Analytics. Write a report on your findings.
3 We want to find some information about the music business in Italy. First, generate some
synonyms using a thesaurus (either a real book or one built into your software). Next,
translate these words from English to Italian using a language translator. Then cut and
paste these words into a search engine and look at the results. Read each of the websites
by selecting the word ‘translate’ at each result.
4 Consult the proposal in the Market Researcher’s Toolbox. Draft a possible design for
secondary data collection. Try to list precisely what you will do. A second task is to carry
this out.
5 In the opening Snapshot, what are the advantages and disadvantages of using the Tesco
Clubcard to obtain data on shopping behaviour? Why might overseas tourists need to be
researched in a different way? What other ‘blind spots’ are there?
Further reading
• McDaniel, C. and Gates, R. (2002) Marketing Research: The Impact of the Internet, 5th edn.
Cincinnati, OH: South Western Thomson Learning.
Standard market research text with a good section on secondary databases and
Internet searches.
• Mort, D. (2006) Sources of Non-official UK Statistics, 6th edn. Aldershot: Gower.
Information of almost 900 sources from over 500 organisations, with telephone
contact numbers and website addresses.
Key web links
• UK statistics http://www.statistics.gov.uk
• European statistics http://www.epp.eurostat.ec.europa.eu
• MR Web sectors http://www.mrweb.com/sectors
Online Resource Centre
www.oxfordtextbooks.co.uk/orc/bradley3e/
Visit the Online Resource Centre that accompanies this book
to access more learning resources on this chapter topic.
Chapter 3 Secondary data
101