Market Research
WEEK 6
LESSON OBJECTIVES
 •List at least five market research suppliers in South Africa
 •Write a market research brief
 •Draft a market research proposal
 •Carry out an evaluation, analysis and synthesising information
 •Understand the accuracy levels of data
 •Create a comprehensive research budget
 •Apply the appropriate time tabling techniques for the research
 •Create a questionnaire and topic guide
 •Explain the purpose of analysis
 •Explain the nature of qualitative analysis
 •Explain the nature of quantitative analysis
 •Describe ways to assess the accuracy of results
Introduction to Market Research
 Market researchers are expected to continuously and intelligently scan the
markets in order to provide businesses with accurate information for efficient
decision making. The challenging part of market research is that it is a costly
and time-consuming process which should be conducted when mega decisions
are involved or where little information is known about the problem (Hague
et al, 2016). A market researcher should be able to communicate and present
complex data. In other words, data is considered to be the market
researcher’s best friend.
 According to Bryman and Bell (2012), research design is “a framework for
collection and analysis of data”. On the other hand, Hague et al, (2016)
defined research design as a process that includes the market research brief,
the market research proposal, evaluation, analysing and synthesising
information, developing the budgeting process and applying appropriate time
tabling techniques for the research.
Market Research Brief
 A research brief or proposal is very vital where the market research is
subcontracted to a research company. The proposal is the instrument that provides
detail to the research company on the needs of its clients. In addition, it provides
an indication of the level of trade-off between delivery time frames, research
depth, the nature of information collected and budgetary constraints (Hague et al,
2016).
 Sudheesh et al (2016), on the other hand, stated that a research proposal consists
of, (1) an introduction, (2) aims and objectives of the research, (3) a review of
literature, (4) research methods and design, (5) ethical considerations, (6)
reference list and (7) Appendices.
 In market research, the business may conduct a research on the impact of pricing
on consumer purchases. In this case, the price is the independent variable and the
consumer purchasing trends are the dependent variable. Hague et al, (2016)
consider a problem statement to be a market research brief or a statement of
opportunity. The brief is the foundation of the market research that determines
the methods that will be used in the research process.
 Hague et al, (2016) insisted that a research brief should provide; possible risks of
the research, causes of the problem, the knowledge of research already done
within the same scope, the target research groups, the specific data the research
should provide, the proposed budget, reporting requirements and the time frames
for delivering research results.
(Problem Statement)
 A problem statement is a clear account of the existing problem to be
investigated, supported by the evidence of the existence of the problem, its
possible causes and the development trends of the problem (Fischler: Online).
 A problem statement is a question or sentence that provide the dependent
variable and the independent variable to be researched on. A dependent
variable is an event being studied that change in response to a change in the
independent variable (Dictionary.com: Online).
RESEARCH PROPOSAL TEMPLATE
Introduction The first part contains a title page, table of contents
and a summary. The introduction should provide a
background the research that prompted the drafting
of the proposal.
The scope The scope of the research stipulates the
geographical area and the target respondents.
Exclusions from the study should be stated here.
The objectives The research objective determines what the
research should focus on.
The methods The research should state the choice between
qualitative and quantitative, primary and secondary
data, telephonic interviews, one on one interviews
and online surveys.
The research team The qualifications and experience of the research
team should be stated.
Timing and costs The timelines and costs of the market research are
stated here.
Evaluation, Analysis and Synthesising
Information
 A market researcher should have detailed knowledge on how to evaluate,
analyse and synthesise information in order to determine the appropriate
information that should be included in the research objectives and what
should be ignored (Hague et al, 2016).
 The information below summarises the process involved in determining the
nature of the information that may be required for the market research.
Outcomes, objectives, and questions Level 1
 Outcome/aim: To launch a new product
 Level 2 Objective of the research
 •To find out the preferred product specifications
 •To establish maximum pricing for the new product
 • To measure customer preferences between the new and old product
 • To determine the possible buying trends in different market segments
 Level 3 Research questions
 •Which product specifications will appeal to the customers?
 •What is the maximum price that can be charged for the new product?
 •Will customers prefer the new product over the existing ones?
 •Which target market is more likely to buy more of the new product?
The Steps in Questionnaire Design
 The steps of questionnaire design one may say would assist in understanding the
problem and finding a solution.
 Steps in questionnaire design:
 1. Formulate hypotheses
 2. Choose collection method
 3. List topics
 4. Plan analysis
 5. Draw a diagram
 6. Lay out the form
 7. Approve and pilot the form
 8. Fine-tune the form.
 Refer to pages 206-209 in your prescribed textbook for a comprehensive
explanation of questionnaire design.
Questionnaire design.
Critical Examination of Marketing Research
and Techniques
Analysis
 Analysis concerns the identification of meaningful patterns in data. If we look at
sales of products from supermarkets and take the outside temperature into
account, we see some fascinating relationships.
 When there is a sudden cold spell, we see that there is an increase in sales of
‘comfort foods’ such as root vegetables and soups. Such analysis has a practical
use: the layout of stores can be modified and stock can be ordered according to
weather forecasts. There are many ways to analyse data, but all of these cannot
be applied in all cases. If data have been collected in a certain way, using certain
instruments, then any analysis will be limited by those constraints.
 Quantitative studies can take full advantage of statistical techniques in attempting
to analyse data. Such studies embrace univariate, bivariate, and multivariate
analysis. There is also a useful tool to assess accuracy, known as significance
testing. Qualitative researchers have different challenges. An eating metaphor is a
good way to visualise this issue. Consider the preparation of food—we cannot make
a fruit dessert without fruit.
 Additionally, the researcher must be alert to the danger of collecting too much
information.
 There is a risk of being unable to see the essential facts when there are so many
facts available.
Data Analysis
 Analysis allows the meaningful patterns to be identified. There are many ways to
analyse data, but if data have been collected in a certain way, using certain
instruments, then analysis will be limited by those constraints. Quantitative
studies can take full advantage of statistical techniques in attempting to analyse
data; qualitative researchers have different challenges. Additionally, the
researcher must be alert to the danger of collecting too much information. There
is a risk of missing the essential points when there are too many facts available.
 In qualitative research, analysis begins before data collection ends and the project
is modified as it progresses. Notes, audio recordings, videos, and transcripts are
analysed by: data reduction, displaying the data, and drawing conclusions.
Different types of qualitative analysis include thematic analysis, radial diagrams,
laddering techniques, explanation building, and the gestalt approach. Radial
diagrams can be a useful and invaluable tool in interview administration and
subsequent analysis. Qualitative analysis software can mean long data preparation
time but programs find concealed data; reanalysis is easy and report-writing is
extremely efficient. Researchers are advised to finish analysis as soon as possible
after fieldwork ends.
Data Analysis
 Coded data are input into a computer manually or scanned. Tables are then produced. Cell or
rim-weighting is used to ensure that the sample is balanced. Grossing means that figures are
multiplied up to the population levels.
 Bivariate analysis takes two variables at a time and inspects the pattern between them and
correlation is the measure of the nature and the strength of association between two
variables.
 Multivariate analysis takes three or more variables at a time and inspects the pattern
between them. Regression is used to analyse associative relationships. Other techniques of
relevance include analysis of variance (ANOVA), factor analysis, discriminant analysis, cluster
analysis, CHAID, and data fusion.
 The accuracy of results depends on many aspects. Sampling error and non-sampling error
account for variations. Non-sampling error may relate to the sampling frame or the process of
data capture.
 Significance testing shows if there is a meaningful difference between the findings from the
sample selected, and what the findings would have been if we had carried out a census. Tests
and assumptions are based on random sampling, but many practitioners use them with
studies that have used non-random sampling. There are several aids available to assess
significance: ready reckoners, nomograms, and software devices.
Analysis of Qualitative Data
 A unique characteristic of qualitative research is that analysis actually begins
before data collection ends. If we take a research project with five group
discussions, analysis begins within the first minutes of the first group and
continues until the last minutes of the final group. Unlike the structured
questionnaire, the focus group agenda requires that the moderator opens up areas
of investigation and probes them until they yield nothing new. In qualitative
approaches, the project is modified as it progresses (Bradley, 2013: 306).
 Questions may therefore be formulated at the interview in response to points
made by respondents. In practical terms, you may be faced with notes from
groups and depth interviews; you may have sound tape recordings, videos, and
perhaps transcripts. Transcripts take the appearance of typed-out versions of
conversations. Historically, researchers would take a pair of scissors and cut them
up, making little piles for each subject that emerged (Bradley, 2013: 306).
 These isolated bits of information could then be subjected to tabular or matrix
analysis. This does not mean statistical tabulations, but one of words and quotes.
The best way to do this is to take an extremely large piece of paper (poster size)
and to divide it into cells.
 For report writing this matrix is used as an aide-memoire; it allows comments to
be extracted and dominant themes to be seen; it may also show the links between
items.
Data Analysis
 In simplistic terms, qualitative data are analysed by:
 1. Data reduction to organise and develop categories
 2. Displaying the data to create text, charts, flow diagrams, matrices
 3. Drawing conclusions to describe patterns and give explanations (Bradley, 2013:
306).
 Several terms are used to describe different types of qualitative analysis. These
include ‘thematic analyses, ‘radial diagrams’, ‘laddering’ techniques, ‘explanation
building’, and the ‘gestalt’ approach. Despite the rather elaborate names, these
methods simply assist the researcher to make sense of a great deal of data and to
answer the research problem.
 In some cases, researchers may do ‘follow-ups’ by consulting participants and use
their transcripts to probe further into what was really meant. This is a form of
triangulation and it is done to confirm findings. Dolan and Ayland (2001) made a
comparison of three approaches to the analysis of qualitative data. These were: 1.
holistic; 2. cut and paste; 3. computer-assisted. The study favours the ‘cut and
paste’ approach (Bradley, 2013: 307).
Data Analysis
 There are important differences between analysis of depth interviews and groups:
the results of individual interviews place different demands on the researcher.
Careful recording and complete transcription may be easier for solo interviews
than for groups, but the bridges between themes are often clearer to see in group
discussions.
 Typically, a client will be debriefed during fieldwork (the client will probably be at
some or all of the groups to receive feedback) and an agreement may be made to
modify the approach. This may mean excluding or including questions or stimulus
materials, and perhaps even replacing a certain profile of respondent with another
(Bradley, 2013: 307).
 During fieldwork, most moderators take notes and write short memos for their
own use.
 This is evidence of early analysis of this complex data set. After the data have
been collected, the researcher is left with recordings and notes. Along with
memories of the sessions and any stimulus materials used, the task of making
sense of the research can be progressed (Bradley, 2013: 307).
Analysis of Qualitative Data
 Consider a study of restaurant diners: the team of researchers, by the end of the
study, has become a team of experts. They can now give their clients different
levels of service, from the most basic to the most ‘deluxe’:
 1. Raw data—questionnaires
 2. Processed data—tables
 3. Analytical report—a commentary on the tables
 4. Interpretative report—a commentary, but with the researcher’s own views
added.
 The quantitative researcher must edit raw data; these data are then coded and
finally input into a computer. If computer-assisted data capture is being used,
these processes may happen simultaneously (Bradley, 2013: 313).
Analysis of Qualitative Data
 ‘Editing’ ensures that answers are complete, accurate, and suitable for further
processing.
 Questionnaires are inspected to identify questions that are answered incorrectly
or not answered at all. This is sometimes called ‘cleaning’ the data, but it is just
one ‘cleaning session’— if the preliminary tables look ‘dirty’, they too will be
cleaned. Tables are inspected to ensure that there are no obvious irregularities;
this is known as ‘sense checking’. The researcher can do various things with any
identified problems:
 • Deduce the answer by inspecting other answers from the same respondent
 • Deduce the answer by inspecting answers from other respondents
 • Return to the respondent and ask the question again
 • Discard or reject the entire data record for the given respondent
 • Do nothing and leave the data ‘dirty’ (Bradley, 2013: 313).
Analysis of Qualitative Data
 Coding’ is the name given to the procedure whereby complex descriptions are
broken into simpler meanings and are allocated a code, usually a number. Closed
questions are generally pre-coded. This means that numbers have already been
allocated to the possible answers. These may have been allocated after a pilot
study, or because the codes are clearly defined (e.g. brand names). Most closed
questions include the ‘other’ category in which other answers appear; these
responses need to be inspected and coded (Bradley, 2013: 314).
 When informants are allowed to answer a question in their own words—expressing
their views freely and without further direction by an interviewer—this is known as
an ‘ open-ended question’. Open-ended questions generate answers that may not
have been anticipated, making them similar to the ‘other’ category (Bradley,
2013: 314).
 In an interview that is administered by an interviewer—for example, face-to-face
or by telephone—the answers will be written down (or typed) verbatim. In a self-
completion interview, the respondent will express views using familiar language
(Bradley, 2013: 314).
Analysis of Qualitative Data
 It is not entirely true to say that the respondent is free to answer such questions.
There are various factors that will influence the length, quality, and content of
response. The length might be determined by the space available to record
comments, and also by the real and perceived time available.
 Similarly, the quality depends on such factors as the time allowed for the question
to be considered, the integrity of the interview situation, and the respondent’s
interest in the topic. The content of responses is also affected by time aspects,
but by the position of the question in the interview: preceding concepts can
actually ‘lead’ the informant towards a specific thought and answer (Bradley,
2013: 314).
 Verbatim answers need to be categorised after interviews have taken place. This
categorisation is commonly described as coding, because numerical codes are
applied to commonly occurring word combinations. This allows questionnaires to
be analysed by computer and quantities of response to be measured (Bradley,
2013: 314).
Analysis of Qualitative Data
 Coding open-ended questions is a time-consuming activity and, where pre-
codes can be created before questionnaires are administered, this is
preferable. The researcher needs to make sense of open-ended questions.
This involves a considered process of extracting and classifying key concepts,
as follows:
 1. A number of answers are written out (often on 50 questionnaires)
 2. Answers that occur frequently are noted (using five-bar tally gates)
 3. Code frames or ‘codebooks’ are created
 4. The ‘code frame’ is used to allocate code numbers to each answer
 5. The code numbers are put into the computer for further analysis (Bradley,
2013: 314).
 Refer to pages 314 to 326 in your prescribed textbook for a comprehensive
explanation of analysis processes.
Data Storage
 Data files are the way in which we store data, where we find it. Before it is ‘fi
led’, it must be in a format that is known and useful, otherwise any future
consultation will be impossible.
 One important aspect of stored data is whether it is ‘multi-punched’ or not. This
term and others go back to Hollerith’s cards. Each column and row was given a
number and the holes were registered. If one column had more than two holes, it
was called ‘multi-punch’. We will look at ASCII, IBM Column Binary, CSV Files, and
ITE Fiche, paying particular attention to the aspect of Multi-punch (Bradley, 2013:
326).
 ASCII is an acronym of American Standard Code for Information Interchange. In
ASCII, there can be no multi-punched data, so the input data needs to
accommodate separate columns for each.
 IBM Column Binary is perhaps the most practical of the data fi les. Each data
position contains 12 binary codes (1 to 9, 0, -, &.) and these can be multi-
punched. CSV files is an acronym of ‘comma-separated variable fi les’; here,
information is separated by commas.
Data Storage
 They can be easily read on spreadsheets such as Excel. Finally, the ITE Fiche
is used to keep electronic table fi les instead of having paper copies. They
take up less space than normal tables. An ITE browser views them (Bradley,
2013: 326).
 These descriptions highlight a problem, particularly for international
research: software vendors and users have not created software that connects
easily. The same questionnaire may be written several times in different
countries using different software. The result is that errors in conversion can
affect a project quality. An organisation called OpenSurvey was created to
promote the concept of open standards in survey software (Bradley, 2013:
326).
Revision Questions
 1. What lessons can we draw from the past to help in analysis today?
 2. In the analysis of qualitative research, transcripts are sometimes not
used. Why is this the case?
 3. What are the main features of quantitative tables?
 4. What are the differences between bivariate and multivariate analysis?
 5. Define these terms using examples: correlation, regression, the punch
card, radial diagrams.
 6. Explain what steps does one take when developing a market
questionnaire?
 7. What is a market brief?

Market Research WEEK 6 LESSON PLAN .pptx

  • 1.
  • 2.
    LESSON OBJECTIVES  •Listat least five market research suppliers in South Africa  •Write a market research brief  •Draft a market research proposal  •Carry out an evaluation, analysis and synthesising information  •Understand the accuracy levels of data  •Create a comprehensive research budget  •Apply the appropriate time tabling techniques for the research  •Create a questionnaire and topic guide  •Explain the purpose of analysis  •Explain the nature of qualitative analysis  •Explain the nature of quantitative analysis  •Describe ways to assess the accuracy of results
  • 3.
    Introduction to MarketResearch  Market researchers are expected to continuously and intelligently scan the markets in order to provide businesses with accurate information for efficient decision making. The challenging part of market research is that it is a costly and time-consuming process which should be conducted when mega decisions are involved or where little information is known about the problem (Hague et al, 2016). A market researcher should be able to communicate and present complex data. In other words, data is considered to be the market researcher’s best friend.  According to Bryman and Bell (2012), research design is “a framework for collection and analysis of data”. On the other hand, Hague et al, (2016) defined research design as a process that includes the market research brief, the market research proposal, evaluation, analysing and synthesising information, developing the budgeting process and applying appropriate time tabling techniques for the research.
  • 4.
    Market Research Brief A research brief or proposal is very vital where the market research is subcontracted to a research company. The proposal is the instrument that provides detail to the research company on the needs of its clients. In addition, it provides an indication of the level of trade-off between delivery time frames, research depth, the nature of information collected and budgetary constraints (Hague et al, 2016).  Sudheesh et al (2016), on the other hand, stated that a research proposal consists of, (1) an introduction, (2) aims and objectives of the research, (3) a review of literature, (4) research methods and design, (5) ethical considerations, (6) reference list and (7) Appendices.  In market research, the business may conduct a research on the impact of pricing on consumer purchases. In this case, the price is the independent variable and the consumer purchasing trends are the dependent variable. Hague et al, (2016) consider a problem statement to be a market research brief or a statement of opportunity. The brief is the foundation of the market research that determines the methods that will be used in the research process.  Hague et al, (2016) insisted that a research brief should provide; possible risks of the research, causes of the problem, the knowledge of research already done within the same scope, the target research groups, the specific data the research should provide, the proposed budget, reporting requirements and the time frames for delivering research results.
  • 5.
    (Problem Statement)  Aproblem statement is a clear account of the existing problem to be investigated, supported by the evidence of the existence of the problem, its possible causes and the development trends of the problem (Fischler: Online).  A problem statement is a question or sentence that provide the dependent variable and the independent variable to be researched on. A dependent variable is an event being studied that change in response to a change in the independent variable (Dictionary.com: Online).
  • 6.
    RESEARCH PROPOSAL TEMPLATE IntroductionThe first part contains a title page, table of contents and a summary. The introduction should provide a background the research that prompted the drafting of the proposal. The scope The scope of the research stipulates the geographical area and the target respondents. Exclusions from the study should be stated here. The objectives The research objective determines what the research should focus on. The methods The research should state the choice between qualitative and quantitative, primary and secondary data, telephonic interviews, one on one interviews and online surveys. The research team The qualifications and experience of the research team should be stated. Timing and costs The timelines and costs of the market research are stated here.
  • 7.
    Evaluation, Analysis andSynthesising Information  A market researcher should have detailed knowledge on how to evaluate, analyse and synthesise information in order to determine the appropriate information that should be included in the research objectives and what should be ignored (Hague et al, 2016).  The information below summarises the process involved in determining the nature of the information that may be required for the market research.
  • 8.
    Outcomes, objectives, andquestions Level 1  Outcome/aim: To launch a new product  Level 2 Objective of the research  •To find out the preferred product specifications  •To establish maximum pricing for the new product  • To measure customer preferences between the new and old product  • To determine the possible buying trends in different market segments  Level 3 Research questions  •Which product specifications will appeal to the customers?  •What is the maximum price that can be charged for the new product?  •Will customers prefer the new product over the existing ones?  •Which target market is more likely to buy more of the new product?
  • 9.
    The Steps inQuestionnaire Design  The steps of questionnaire design one may say would assist in understanding the problem and finding a solution.  Steps in questionnaire design:  1. Formulate hypotheses  2. Choose collection method  3. List topics  4. Plan analysis  5. Draw a diagram  6. Lay out the form  7. Approve and pilot the form  8. Fine-tune the form.  Refer to pages 206-209 in your prescribed textbook for a comprehensive explanation of questionnaire design.
  • 10.
  • 11.
    Critical Examination ofMarketing Research and Techniques Analysis  Analysis concerns the identification of meaningful patterns in data. If we look at sales of products from supermarkets and take the outside temperature into account, we see some fascinating relationships.  When there is a sudden cold spell, we see that there is an increase in sales of ‘comfort foods’ such as root vegetables and soups. Such analysis has a practical use: the layout of stores can be modified and stock can be ordered according to weather forecasts. There are many ways to analyse data, but all of these cannot be applied in all cases. If data have been collected in a certain way, using certain instruments, then any analysis will be limited by those constraints.  Quantitative studies can take full advantage of statistical techniques in attempting to analyse data. Such studies embrace univariate, bivariate, and multivariate analysis. There is also a useful tool to assess accuracy, known as significance testing. Qualitative researchers have different challenges. An eating metaphor is a good way to visualise this issue. Consider the preparation of food—we cannot make a fruit dessert without fruit.  Additionally, the researcher must be alert to the danger of collecting too much information.  There is a risk of being unable to see the essential facts when there are so many facts available.
  • 12.
    Data Analysis  Analysisallows the meaningful patterns to be identified. There are many ways to analyse data, but if data have been collected in a certain way, using certain instruments, then analysis will be limited by those constraints. Quantitative studies can take full advantage of statistical techniques in attempting to analyse data; qualitative researchers have different challenges. Additionally, the researcher must be alert to the danger of collecting too much information. There is a risk of missing the essential points when there are too many facts available.  In qualitative research, analysis begins before data collection ends and the project is modified as it progresses. Notes, audio recordings, videos, and transcripts are analysed by: data reduction, displaying the data, and drawing conclusions. Different types of qualitative analysis include thematic analysis, radial diagrams, laddering techniques, explanation building, and the gestalt approach. Radial diagrams can be a useful and invaluable tool in interview administration and subsequent analysis. Qualitative analysis software can mean long data preparation time but programs find concealed data; reanalysis is easy and report-writing is extremely efficient. Researchers are advised to finish analysis as soon as possible after fieldwork ends.
  • 13.
    Data Analysis  Codeddata are input into a computer manually or scanned. Tables are then produced. Cell or rim-weighting is used to ensure that the sample is balanced. Grossing means that figures are multiplied up to the population levels.  Bivariate analysis takes two variables at a time and inspects the pattern between them and correlation is the measure of the nature and the strength of association between two variables.  Multivariate analysis takes three or more variables at a time and inspects the pattern between them. Regression is used to analyse associative relationships. Other techniques of relevance include analysis of variance (ANOVA), factor analysis, discriminant analysis, cluster analysis, CHAID, and data fusion.  The accuracy of results depends on many aspects. Sampling error and non-sampling error account for variations. Non-sampling error may relate to the sampling frame or the process of data capture.  Significance testing shows if there is a meaningful difference between the findings from the sample selected, and what the findings would have been if we had carried out a census. Tests and assumptions are based on random sampling, but many practitioners use them with studies that have used non-random sampling. There are several aids available to assess significance: ready reckoners, nomograms, and software devices.
  • 14.
    Analysis of QualitativeData  A unique characteristic of qualitative research is that analysis actually begins before data collection ends. If we take a research project with five group discussions, analysis begins within the first minutes of the first group and continues until the last minutes of the final group. Unlike the structured questionnaire, the focus group agenda requires that the moderator opens up areas of investigation and probes them until they yield nothing new. In qualitative approaches, the project is modified as it progresses (Bradley, 2013: 306).  Questions may therefore be formulated at the interview in response to points made by respondents. In practical terms, you may be faced with notes from groups and depth interviews; you may have sound tape recordings, videos, and perhaps transcripts. Transcripts take the appearance of typed-out versions of conversations. Historically, researchers would take a pair of scissors and cut them up, making little piles for each subject that emerged (Bradley, 2013: 306).  These isolated bits of information could then be subjected to tabular or matrix analysis. This does not mean statistical tabulations, but one of words and quotes. The best way to do this is to take an extremely large piece of paper (poster size) and to divide it into cells.  For report writing this matrix is used as an aide-memoire; it allows comments to be extracted and dominant themes to be seen; it may also show the links between items.
  • 15.
    Data Analysis  Insimplistic terms, qualitative data are analysed by:  1. Data reduction to organise and develop categories  2. Displaying the data to create text, charts, flow diagrams, matrices  3. Drawing conclusions to describe patterns and give explanations (Bradley, 2013: 306).  Several terms are used to describe different types of qualitative analysis. These include ‘thematic analyses, ‘radial diagrams’, ‘laddering’ techniques, ‘explanation building’, and the ‘gestalt’ approach. Despite the rather elaborate names, these methods simply assist the researcher to make sense of a great deal of data and to answer the research problem.  In some cases, researchers may do ‘follow-ups’ by consulting participants and use their transcripts to probe further into what was really meant. This is a form of triangulation and it is done to confirm findings. Dolan and Ayland (2001) made a comparison of three approaches to the analysis of qualitative data. These were: 1. holistic; 2. cut and paste; 3. computer-assisted. The study favours the ‘cut and paste’ approach (Bradley, 2013: 307).
  • 16.
    Data Analysis  Thereare important differences between analysis of depth interviews and groups: the results of individual interviews place different demands on the researcher. Careful recording and complete transcription may be easier for solo interviews than for groups, but the bridges between themes are often clearer to see in group discussions.  Typically, a client will be debriefed during fieldwork (the client will probably be at some or all of the groups to receive feedback) and an agreement may be made to modify the approach. This may mean excluding or including questions or stimulus materials, and perhaps even replacing a certain profile of respondent with another (Bradley, 2013: 307).  During fieldwork, most moderators take notes and write short memos for their own use.  This is evidence of early analysis of this complex data set. After the data have been collected, the researcher is left with recordings and notes. Along with memories of the sessions and any stimulus materials used, the task of making sense of the research can be progressed (Bradley, 2013: 307).
  • 17.
    Analysis of QualitativeData  Consider a study of restaurant diners: the team of researchers, by the end of the study, has become a team of experts. They can now give their clients different levels of service, from the most basic to the most ‘deluxe’:  1. Raw data—questionnaires  2. Processed data—tables  3. Analytical report—a commentary on the tables  4. Interpretative report—a commentary, but with the researcher’s own views added.  The quantitative researcher must edit raw data; these data are then coded and finally input into a computer. If computer-assisted data capture is being used, these processes may happen simultaneously (Bradley, 2013: 313).
  • 18.
    Analysis of QualitativeData  ‘Editing’ ensures that answers are complete, accurate, and suitable for further processing.  Questionnaires are inspected to identify questions that are answered incorrectly or not answered at all. This is sometimes called ‘cleaning’ the data, but it is just one ‘cleaning session’— if the preliminary tables look ‘dirty’, they too will be cleaned. Tables are inspected to ensure that there are no obvious irregularities; this is known as ‘sense checking’. The researcher can do various things with any identified problems:  • Deduce the answer by inspecting other answers from the same respondent  • Deduce the answer by inspecting answers from other respondents  • Return to the respondent and ask the question again  • Discard or reject the entire data record for the given respondent  • Do nothing and leave the data ‘dirty’ (Bradley, 2013: 313).
  • 19.
    Analysis of QualitativeData  Coding’ is the name given to the procedure whereby complex descriptions are broken into simpler meanings and are allocated a code, usually a number. Closed questions are generally pre-coded. This means that numbers have already been allocated to the possible answers. These may have been allocated after a pilot study, or because the codes are clearly defined (e.g. brand names). Most closed questions include the ‘other’ category in which other answers appear; these responses need to be inspected and coded (Bradley, 2013: 314).  When informants are allowed to answer a question in their own words—expressing their views freely and without further direction by an interviewer—this is known as an ‘ open-ended question’. Open-ended questions generate answers that may not have been anticipated, making them similar to the ‘other’ category (Bradley, 2013: 314).  In an interview that is administered by an interviewer—for example, face-to-face or by telephone—the answers will be written down (or typed) verbatim. In a self- completion interview, the respondent will express views using familiar language (Bradley, 2013: 314).
  • 20.
    Analysis of QualitativeData  It is not entirely true to say that the respondent is free to answer such questions. There are various factors that will influence the length, quality, and content of response. The length might be determined by the space available to record comments, and also by the real and perceived time available.  Similarly, the quality depends on such factors as the time allowed for the question to be considered, the integrity of the interview situation, and the respondent’s interest in the topic. The content of responses is also affected by time aspects, but by the position of the question in the interview: preceding concepts can actually ‘lead’ the informant towards a specific thought and answer (Bradley, 2013: 314).  Verbatim answers need to be categorised after interviews have taken place. This categorisation is commonly described as coding, because numerical codes are applied to commonly occurring word combinations. This allows questionnaires to be analysed by computer and quantities of response to be measured (Bradley, 2013: 314).
  • 21.
    Analysis of QualitativeData  Coding open-ended questions is a time-consuming activity and, where pre- codes can be created before questionnaires are administered, this is preferable. The researcher needs to make sense of open-ended questions. This involves a considered process of extracting and classifying key concepts, as follows:  1. A number of answers are written out (often on 50 questionnaires)  2. Answers that occur frequently are noted (using five-bar tally gates)  3. Code frames or ‘codebooks’ are created  4. The ‘code frame’ is used to allocate code numbers to each answer  5. The code numbers are put into the computer for further analysis (Bradley, 2013: 314).  Refer to pages 314 to 326 in your prescribed textbook for a comprehensive explanation of analysis processes.
  • 22.
    Data Storage  Datafiles are the way in which we store data, where we find it. Before it is ‘fi led’, it must be in a format that is known and useful, otherwise any future consultation will be impossible.  One important aspect of stored data is whether it is ‘multi-punched’ or not. This term and others go back to Hollerith’s cards. Each column and row was given a number and the holes were registered. If one column had more than two holes, it was called ‘multi-punch’. We will look at ASCII, IBM Column Binary, CSV Files, and ITE Fiche, paying particular attention to the aspect of Multi-punch (Bradley, 2013: 326).  ASCII is an acronym of American Standard Code for Information Interchange. In ASCII, there can be no multi-punched data, so the input data needs to accommodate separate columns for each.  IBM Column Binary is perhaps the most practical of the data fi les. Each data position contains 12 binary codes (1 to 9, 0, -, &.) and these can be multi- punched. CSV files is an acronym of ‘comma-separated variable fi les’; here, information is separated by commas.
  • 23.
    Data Storage  Theycan be easily read on spreadsheets such as Excel. Finally, the ITE Fiche is used to keep electronic table fi les instead of having paper copies. They take up less space than normal tables. An ITE browser views them (Bradley, 2013: 326).  These descriptions highlight a problem, particularly for international research: software vendors and users have not created software that connects easily. The same questionnaire may be written several times in different countries using different software. The result is that errors in conversion can affect a project quality. An organisation called OpenSurvey was created to promote the concept of open standards in survey software (Bradley, 2013: 326).
  • 24.
    Revision Questions  1.What lessons can we draw from the past to help in analysis today?  2. In the analysis of qualitative research, transcripts are sometimes not used. Why is this the case?  3. What are the main features of quantitative tables?  4. What are the differences between bivariate and multivariate analysis?  5. Define these terms using examples: correlation, regression, the punch card, radial diagrams.  6. Explain what steps does one take when developing a market questionnaire?  7. What is a market brief?