3. Learning Outcomes
• To describe and understand qualitative
data gathering tools
• To describe and understand quantitative
data gathering tools
• Overview methods of data analysis
• Lectures comes from: Thomas, G. (2009)
how to do your research project. London.
Sage. Chapter 8
4. 1. Introduction
• Once you have decided how you are going to
approach your question and the broad design
frame that will be used, you now need to
decide how you are going to collect your data.
• This means the different instruments and
techniques with which you will gather
information.
• Do not come up with the tool first and then
find a way of using it, because if the only tool
you have is a hammer you will treat
everything as if it were a nail.
• Do not let your method dominate your
research process.
5. 2. Method: the way of doing
something systematically
• Method is a systematic structured approach to
gathering data.
• It is a considered thought through way of
approaching your research question in order to
find the answer you are seeking.
• Some of these methods collect data many
comprising words (qualitative research);
• some convert information into numbers
(qualitative research);
• some collect both words and numbers (mixed
method research).
6. 2. Method: the way of doing
something systematically
• Tools, methods, techniques and instruments are terms that
are often used interchangeably, resulting in confusion.
• Screwdrivers, chisels and hammers are undeniably tools -
not methods.
• But they have to be used with a method: watching an
inexperienced person wields a chisel is a painful
experience. You need to know how to use the chisel. Until
you have learnt the method, the chisel is as good as
useless.
• So tool and method go together: hand in glove!
• The method is almost the tool in itself which is why the terms
are often seen as synonymous and used together.
• However it is important to keep the concept of method
separate from the instruments of data collection.
7. 2. Method: the way of doing
something systematically
• The instruments that you will use in data
collection will depend on the type of data
being gathered: qualitative and/or
quantitative data.
• Whilst different methods have conventions
rules and procedures, the different
instruments can be used creatively.
• Using the chisel example, a key method in
woodcarving is not to of against the grain,
however the chisel can create a thing of
beauty when used creatively.
8. 3. QUALITATIVE DATA
GATHERING TOOLS
1. Interviews
2. Accounts
3. Diaries
4. Group interviews and focus groups
5. Document interrogation
9. 3. Qualitative data gathering tools
• Reflect on the similarities and differences between
qualitative research and therapy..... Like therapists, the
researcher must choose between competing practices
and theoretical traditions. In their attempt to describe
and interpret personal experience, the researcher
needs to have an open mind about where their
research journey will take them.
• Qualitative researchers want to explore peoples stories.
The focus is on attempting to make sense of
phenomena in terms of the social meanings people
bring.
• Qualitative research begins not with hypothesis to be
tested or causal relationships to be established but
rather with open research questions:
10. 3. Qualitative data gathering tools
• Qualitative research cannot answer the
question such as: "what women develop eating
disorders?" But rather it might ask "how do
women with anorexia make sense of why they
have developed the condition?"
• Rich, textured description is valued along with
focus on the "how's" and the "what's" rather
than the "why" and the "how many". The
research questions like "how mental health
problems represented in the media?" Or "what
is it like to experience a traumatic relationship
breakup?".
11. 3. Qualitative data gathering tools
• It is important to avoid questions which contain an implicit
hypothesis. For example "what the perceived benefits to
victims of domestic violence of self-help groups?" Contains
the assumption that such groups or helpful it is best to
have a narrow focused open research question: "how do
victims of domestic violence experienced self-help
groups?“
• Qualitative researchers understand that the world cannot
be understood in clear-cut cause-and-effect terms.
complexity and ambivalence par for the course.
• The researchers own role in the research context are
understood to be part of the complexity. The researcher
recognises they are part of what is being studied, and
acknowledge the impact on the research through
reflexivity.
12. 3. Qualitative data gathering tools
• Counsellors are drawn to qualitative research
because it similar to therapy and resonates with
us.
• Both are concerned with mutual discovery,
exploring meanings and understanding how the
world is experienced by another.
• Both involve a relational process, that promotes
collaborative empowering relationships.
• Familiar skills of interviewing and empathic
listening transferable to the research arena.
• One of the main differences however is that
research aim is to produce knowledge rather
than enable individual awareness or change.
14. 3.1. Interviews
• An interview is a discussion with someone in which
you try to get information from them. The
information may be facts or opinions or attitudes or
experiences or any combination of these.
• The three basic sub types of interview: structured
interviews, unstructured interviews, semi-structured
interviews.
• Interviews involve personal contact either directly
or via the telephone.
• This has a profound effect in the where
interviewees will respond to you in comparison to
how they would have reacted to a questionnaire
coming to the post.
15. 3.1. Interviews
• Because of the primacy of the personal contact, your
appearance, demeanour and tone are important: how
do you want to be seen? As "one of us" or as a
neutral observer or as a person in authority? Your
decision should influence the way you look sound
and behave.
• It is important to establish rapport with your
interviewee at the beginning, before the interview
proper begins. Discuss some neutral topic example
the weather, the journey, etc. It is important in the
process of making contact and establishing grounds
for the interview to begin. With some clients you may
not establish a meaningful rapport.
• You need to ask yourself before the interviews begin
what it is that you are trying to get from your
interviewees and how the personal contact will help.
16. 3.1. Interviews
• Does your design mean that you will be interpreting
what your respondents say, or does it need you to want
to gather straightforward "facts".
• In gathering interpretive data you will be reading your
interviewee's behaviour, mannerisms and gestures as
carefully as the words as these inform you what the
interviewees really means beyond the actual words
they are using.
• Words do change the meaning depending on context,
and meaning goes beyond the words, so we often need
to read into what the other person is saying. How does
this impact on your research data gathering process?
17. 3.1. Interviews
• The written transcripts of the original spoken
words does not pick up the behavioural cues that
you will experience in the interview.
• How will you record these important nuances?
Will you take notes there and then (or very soon
afterwards) all you add them to the audio
recording subsequently? How accurate with this
be?
• It is important to have an accurate record of the
interview if you are doing interpretive research.
• You will need to explain your recording methods
briefly to the interviewee and what is being done
with the data, how it is being stored, analysed
and subsequently destroyed.
18. 3.1.1.Structured interviews
• A structured interview is a meeting with another person in
which you ask a predetermined set of questions. Beyond the set
of questions there is very little scope for further following up.
• The idea behind the structure is that there is a degree of
uniformity provided across different interviewees you meet. The
interviewees responses will be recorded on a form that will
probably be a mix different kinds of response, both open-ended
and closed.
• Open-ended questions allow the respondents to reply in
whatever way they wish: "what are your feelings about the
national lottery?“
• Closed questions of those that demand a particular response:
"do you approve of the national lottery? Yes or no"; or "how
comfortable are you feel about the national lottery? Very
comfortable, comfortable, no opinion, not comfortable, very
uncomfortable."
19. 3.1.1.Structured interviews
• Strengths of a structured interview:
• 1. Relatively easily administered
• 2. Interviewee's responses can be quite easily coded
• The disadvantage of the structured interview is a to
much structure loses the key purpose of the face-to-
face interview, go beyond the mere tick in a box and
get something other than the assured response.
• If you merely achieved ticks in a box you might as
well give a questionnaire.
• Do not be too rigid and lose the key value of the
interview. Allow a degree of flexibility in exploring
their responses to questions in order to gather the
richness of the answer.
20. 3.1.2.Unstructured interviews
• An unstructured interview is like a conversation.
There is no predetermined format to the interview
beyond your topic of interest.
• There is no pre-defined list of questions and no set
agenda.
• The interviewee is allowed to set the agenda,
determining the important issues, allowing them to
tell you what the issues are, what is important to
them.
• As the researcher you go in with an open mind and is
important that the frame set for the research allows
the interviewee the scope to do this.
21. 3.1.2.Unstructured interviews
• Just how "unstructured" is the unstructured
interview?
• If your respondent goes completely off topic then you
might wish to bring them back to it in a careful and
sensitive manner.
• You need to understand the purpose of "off topic"
discussion and how it relates to the topic of interest.
• You would need to prompt interviewee without
setting an agenda to bring them back to topic.
• You could say something like "can you tell me more
about that?" Or "what happened next" but avoid
interpretive questions that might be leading for
example "does that make you feel really angry?".
• Avoid putting words in interviewee's mouth.
22. 3.1.3.Semi-structured interviews
• The semi-structured interview provides the best of both
worlds combining structure of a list of issues to be covered,
with the freedom to follow up points as necessary.
• It is the most common arrangement in most small-scale
research interviews.
• Do not consider using this simply because it is easier as it
might influence the data are you collecting, leading to a
different kind of research from which you have set out to do.
• For example if you are really interested in interpreting your
interviewee's comments and you are a participant observer in
the situation you are researching, an unstructured interview
remains the best choice.
• In order to get the best out of the semistructured interview,
you will need an interview schedule rather than a set of
interview questions.
23. 3.1.3.Semi-structured interviews
• This is a list of issues which you want to cover but do not have to be in
the form of questions but rather act as an aide memoir of the important
points for discussion.
• You do not need to go to the points in order, or keep in any formal way
to the structure but rather these are a reminder of what you intend to
cover.
• Your interview schedule, drawn up prior to the interview, is a framework
of issues, leading to possible questions, leading to possible follow up
questions, leading to probes.
• Probes are encouragements to interviewees from the interviewer to
proceed with aspects of the answers. These are both the verbal
prompts "go on..." or non-verbal prompts for example a nod, a wave of
a hand to encourage further discussion.
• This schedule is a structure to help you conduct the interview. You
should feel free to ask different questions or supplementary questions
as the need arises.
24. 3.2.Accounts
• Accounts are really the products of unstructured
interviews but without the expectation that these will
have been affected by an interview.
• The accounts could for example, depending on your
informant, have been provided in the form of a long
written piece of prose like an essay;
• or it could be an account in an audio form for
subsequent transcription.
• An account will be handled in the same way as the
data from an unstructured interview.
25. 3.3.Diaries
• The diary is an invaluable data gathering tool for the
researcher undertaking a small project.
• A diary is a regular, usually day by day, record of thoughts or
occurrences about events and experiences. It may involve the
participant in your research making a record of thoughts,
feelings, actions responses etc, or it may involve a more
structured record being taken of specific activities.
• The advantage of this process is that some people find the
more personal and private nature of the process allows them
to give more detailed information than they would in a face-
to-face interview.
• Diaries can take different formats: written, audio recordings,
photographic recordings, video recordings. You need to
consider the advantages and disadvantages of each diary
format.
26. 3.3.Diaries
• Diaries can be more than a simple record of what
happened, often it is a record not only of the events but
also the person's interpretations of those events.
• These particular interpretations determined by the
clients context, background, culture et cetera.
• A structured diary can collect data about specific events
and activities relevant to your topic. This results in
easier coding from a variety of participants to ensure
uniformity of data.
• What diaries provide is a longitudinal and regular
collection of data that interview cannot achieve on its
own.
27. 3.4.Group interviews and focus
groups
• Interviewing in a group has particular elements
that you need to be aware of: people behave
differently in groups and the particular ways a
whole group will behave differently from
individuals.
• In a group particular individuals may be more
talkative order less talkative; some people may
take the lead or others follow.
• Also the group may display "risk shift
phenomenon": the group will make a riskier
decision on an individual. Groups tend to
influence the overall information process than
what you would get from individuals. The group
safety plays a vital role in influencing the
information you will gather from a group.
28. 3.4.Group interviews and focus
groups
• You must therefore be aware that you will obtain
different responses from the group, than you would
have obtained from the same people interviewed
individually.
• You need to establish why you are doing a group
interview rather than a set of individual interviews.
• One of the most important reasons for wanting a group
interview would be those concerning group psychology
itself. You want to find out how a group as a whole
behave in relation to a particular event, or an attitude
that the group may hold as a whole.
• In a group interview the research takes on the role
asking questions, and is in control of the discussion.
• this is a discussion between the research and the
participants.
29. 3.4.Group interviews and focus
groups
• A focus group the role of the group leader is more of
a facilitator or moderator. The aim of focus group is to
facilitate discussion among participants are not
between yourself and the participants.as a facilitator
your role is to stimulate discussion through
comments, a range of focused materials and
prompts.
• As groups require facilitation it is difficult to be both
facilitator and recorder of information. It is common
practice to use an observer to record information
about context, environment and participants
behaviour in group settings. It may be helpful to
record proceedings using audio and or video.
30. 3.5.Document interrogation
• Gathering data from documents requires an entirely different
process from gathering data from people. It is important to
find the right documents, to read them in a structured
manner and to analyse them meaningfully. Different
documents require different documentary interrogation in
order to find the data are you seek.
• An example of data documentation analysis could be
something like BACP professional conduct procedure
outcomes. You gather the last 10 years of boards outcomes
and to analyse the decision-making process, and outcomes
looking for changes over time.you would have a clear
structured checklist in which you are looking for particular
themes and ideas while you interrogate the documents.
31. 4.MIXED METHODS: DATA
GATHERING FOR WORDS AND
NUMBERS
1. Questionnaires
2. Scales
Mixed methods involve
tools that collect words or
numbers or both, or they
may commonly convert
the words into numbers.
32. 4.1.Questionnaires
• Questionnaires: the defining characteristic of a
questionnaire is that it is a written form of questioning.
• Questionnaires can be used to collect an array of data
using both open or closed questions; collecting facts,
or attitudes; all be part of a procedure to assess
something in particular for example personality.
• Questionnaires can be presented in a variety of
formats and manner: it can be tightly structured, or
allow the opportunity for more open and discussion of
responses.
• They can be read out the interviewers, or sent to
respondents to complete themselves, they may be
sent by post, email or even be an online questionnaire
(Google Docs, survey monkey).
33. 4.1.Questionnaires
Basic considerations in constructing a questionnaire
• 1. Keep everything short. Try to limit your questionnaire to one side
of A4. Keep the questions short.
• 2. Be clear about what you are asking. Only ask for one piece of
information at a time. Do not ask for two pieces of information in
one sentence, this will confuse respondents.
• 3. Be precise about what you are asking. Give them a choice of
options rather than an open-ended answer
• 4. Collect all the necessary details. Some obvious information that
may help with data analysis data on, needs to be on the form as
you cannot gather it later. Example gender, years of experience et
cetera
• 5. Be aware of "prestige bias": makes is want to appear with all of
the things they can need to prestige (to appear clever, rich,
educated et cetera). Be aware of this in the way you pose
questions and interpret responses. This can lead to the respondent
assuming there is "a right answer".
34. 4.1.Questionnaires
• Kinds of questions and kinds of responses
• Open questions
• Open questions are similar to unstructured interviews
with the same considerations needed. However you
are unable to prompt the respondent in a
questionnaire with "anything else you would like to
say". Rather than asking respondents an open
question at the end of the questionnaire: "is anything
asked you would like to add" which often results in
the syndrome of mind emptying: suddenly you have
no idea what to put in that box.
• Rather structure that open-ended question into a
series of prompts and will cover the topic you are
attempting to answer.
35. 4.1.Questionnaires
• Closed questions
• Close questions can be organised in a number of ways:
• 1. Dichotomous questions: these are usually "either-or" ;
"yes - no" answers. You have a single choice to make
between two options. Is often used as screening
questions in which you can then separate respondents
into different groups.
• 2. Multiple choice questions: contain two or more answers
where respondents can be told either to tick one box to
tick as many boxes as needed. Depending on the purpose
of the multiple choice question will change depending on
the data needed. If you are interested in respondents
knowledge rather than their beliefs, there might only be
one right answer; or multiple classes depending on beliefs
being covered.
36. 4.1.Questionnaires
• 3. Rank order questions: respondents have
to rank items (put them in order) on a list
according to some criterion (best – to –
worst), in which you can ask either limited
choices or require them to rank the whole
list.
• 4. Rating scale questions: require the
respondent to rate some experience,
attribute, attitude along a continuum: very
positive, positive, neutral, negative, very
negative. The respondents will tick only
one of these boxes.
37. 4.1.Questionnaires
• 5. Constant sum method: requires the respondent to
distribute points, usually 100, to a set of answers. You
provide them with a taxonomy (an arrangement of ideas)
contains a number of features associated with the concept
you are exploring and ask them to distribute the points
amongst these features. One of the advantages of this
method is the attribution of a strength of feeling to various
answers revealing the relative importance attributed to
different opinions. This allows statistical manipulation of
the data that would not be possible with other
questionnaires.
• 6. Matrix or grid questions: this provides a series of
questions which all have the same answer scale: example
all on the same scale of 1 to 5. It is important to make
clear to respondents how the scale operates for example
adding the words high or low at each end, or an arrow
indicating level of importance increasing across the scale.
38. 4.2.Scales
• Scales are a set of items and responses that allow for a degree of
measurement. It is not uncommon for scales to be included in
questionnaires.
• The two main scale measurements that are used: . Likert scale & Semantic
differential scale
• 1. Likert scale: primarily used for measuring attitudes.
• Respondent indicate their level of agreement to statements provided by
the research relating to the attitude, belief or characteristic.
• The respondent response to each item on the five point scale usually with
answers from strongly agree, agree, neither agree nor disagree, disagree,
strongly disagree. With the tendency for some people over choose the
middle option, neither agree nor disagree, this middle option is sometimes
removed.
• A Likert scale can be used in any situation where belief or attitude is to be
measured.
• The important thing to remember is that you are asking for agreement or
disagreement with a statement that you provide. It is important that your
statement is clear and unambiguous.
39. 4.2.Scales
• 2. Semantic differential scale: using opposite meaning adjectives,
such as "kind/cruel" or "exciting/boring", the scale requires the
respondents to rate something on a seven point scale in relation to
those adjectives.
• Exciting……………………..boring
• Kind …………………………cruel
• Generous……………………tight
• You use the semantic differential scale to draw a more textured
picture of respondents thinking and look at interesting differences
where they occur between subgroups within your sample. If you do
not intend to be present in the questionnaire you need to provide
an example of an already completed (but irrelevant one) to explain
what needs to be done.
40. 5. OBSERVATIONS
1. Structured observation
2. Unstructured observation
Observation is one of the
most important ways of
collecting data in social
research. Observing means
watching carefully, watching
in some very different ways,
depending on the purpose of
the research. There are two
kinds of observations:
structured observation, and
unstructured observations
41. 5.1.Structured observation
• In structured observation you are making the assumption
that the social world can be broken down into quantifiable
elements, bits of data that you can count. The first thing that
the observer has to do is to define what these bits are to be.
These may be individual pieces of action that occurs, or use
of particular language. The observer has to devise a way of
counting these elements.This is non-participants observation
• 4 Ways of counting in observations:
1. Duration recording: the Observer measures the overall time
that a target behaviour occurs in a particular time period
2. Frequency count according: the Observer records each time
the topic behaviour occurs in a particular time period
42. 5.1.Structured observation
• 3. Interval recording: you decide on an
interval (three seconds, 10 seconds, 20
seconds depending on the complexity of what
you're looking for); target individuals; and
categories of behaviour (on-task, off task). You
will have data which can be processed in a
number of ways, including numerical analysis.
• 4.Time sampling: refers to the fact that you are
selecting intervals after the total time available
for observation and then only observed during
the selected periods. This is used in
conjunction with the three elements above. This
is used in particular for gathering information of
classroom activity.
43. 5.2.Unstructured observation
• Unstructured observation is undertaken when you
are immersing yourself in a social situation, usually
as some kind of participant, in order to understand
what is going on there.
• This kind of observation is often called participant
observation, because it is associated with research is
becoming a participant in the situations they are
researching.
• It entails talking to people, watching, reading
documents, keep units that enable you to
understand the situation. It is more than simple
observation. It is difficult to disentangle where one
kind of participation begins and another ends.
45. Quantitative data gathering tools
• Quantitative data gathering is a process by
which numbers come to represent the
experience of the person. The use of numbers
as a clean conveyor of "truth" is associated
with the positivist paradigms, where knowledge
about the world can be obtained objectively. As
a result it is seen as the "absolute" truth
because it is measurable. This is merely the
positivist paradigms point of view.
• But this notion of clean, simple efficiency in the
transport of knowledge is misleading, for in
social research numbers are only as reliable as
a concepts that underlie them eg IQ scores.
46. 6.1. Measurements and tests
• The use of measurements and tests is a process whereby you
are checking the extent of something.
• The results of a test will nearly always be in a numerical form.
• In social sciences they take varied forms from being formal, or
informal measurements of some attribute, personal feature,
or attainment.
• These tend to be associated with complex well standardised
forms.
• Test construction and standardisation is a large and separate
field of study, beyond the capacities and scope of most
undergraduate dissertations.
• Trying to create your own concept test is what is discouraged,
not questionnaires!
47. 6.1. Measurements and tests
• Tests can be divided into:
• 1.norm referenced: compares the person being tested just
sample of their peers.
• this kind of test aims to compare individuals one against the
other: eg .intelligence tests.
• Standardisation is an important element in the construction of
Norm referenced tests.
• This involves constructing the test under particular specific,
repeatable conditions with large samples from a population.
• a good test is one that is both reliable (refers to the tests
ability to measure something consistently) and valid (is a
measure of how well it is assessing what it is supposed to
measure).
• 2.criterion referenced: assesses whether someone is able
to meet some criteria, irrespective of how well other people
perform on the test.this test merely compares the individual
against the criteria: eg Driving test.
48. 6.2. Official statistics
Official statistics can form the basis of a good research
project, or complement your project, drawing on
relevant statistics.
A wide range of statistics gathered by the Organisation
for Economic Cooperation and Development (OECD)
are available online.
Office for National statistics is another source of data
College statistics can be obtained through a request to
the Administration in the front office of the particular
statistics you are seeking. if this data is available they
will be obtained, but could take some time.
50. 7.Overview of data analysis
methods
• Following the collection of your data you need
to analyse it.
• Your data may be in any variety of forms, and
the method for analysing the data will vary
accordingly.
• There is a wide array of analytical methods for
handling the data you have gathered.
• You need to ensure the importance of the
coherence of your story of your research.
• Your analysis should fit the approach you've
taken in your research.
51. 7.Overview of data analysis
methods
• The most commonly used methods for analysing words:
• 1.constant comparative method;
• 2.network analysis;
• 3.construct mapping and theme mapping;
• 4.grounded theory;
• 5.discourse and content analysis
• The most commonly used methods for analysing numbers:
• 1. Statistics that describe
• 2. Statistics this understand relationship between variables
• 3. Statistics that help deduce or infer
• This will provide a simple overview of these methods. More
detailed lecture on data analysis will follow
52. 7.1. Analysing words
When you have gathered data in words, you're seeking
to use those words in an illuminative analysis of the
situation in which you are interested.
When working with words be aware that you are
working from a interpretivism paradigm:
• Knowledge is everywhere and is socially
constructed
• All kinds of information valid and worthy of the
name "knowledge", even things "of the mind“
• Specific accounts inform each other
• The act of trying to know should be conducted such
that the knower's own value position is taken into
account in the process.
53. interpretivism
• The main point about interpretivism is
that we are interested in people and the
way they interrelate, what they think and
how they form ideas about the world, how
their worlds are constructed.
• The key is understanding. What
understandings do the people we are
talking to have about the world, and how
can we in turn understand these?
54. 7.1.1.Constant comparative
method
• The basic analytic method of the interpretive researcher is
constant comparison.
• This involves going to your data again and again - this is the
constant bit
• Comparing each element - phrase, sentence or paragraph- with
all the other elements - this is the comparative bit
• There is nothing more complicated than that, although there
may be many different ways of going about the comparison.
• From the constant comparison you merge with themes that
capture or summarise the content of your data.
• These themes or categories of the essential building blocks of
your analysis
• There are various ways in which you map your themes to show
the interconnections between them. The two methods used for
mapping themes are network analysis and construct mapping
55. 7.1.2.Network analysis
Network analysis: you and to show how one
idea is related to another using a network,
which is a bit like a tree, with a trunk which is
the basic idea, and branches coming off the
trunk representing constituent ideas.
This is useful where there is a core theme,
which comprises a range of subthemes.
Network analysis shows how the themes
related to one another with each branch
holding a range of ideas
It provides a hierarchical organisation of ideas
contained in your data.
56. 7.1.3.Constructed mapping and
theme mapping
Construct mapping quits themes into sequential order from the
interview and uses lines and arrows to make connections between
the ideas and themes. It developed out of the idea of George
Kelly's personal construct theory. This is a complex theoretical
lens in which to analyse data.
In a similar manner, theme mapping, using constant comparative
method, helps establish the themes.
Once you've established themes, you go through your working
data files and look for good quotations that illustrate those
themes.
Then,in order that those quotations appear in the interview, put
them into boxes on the page. The page now becomes your map.
Enable the boxes with the names of the themes and draw dotted
lines if they seem to be connected and solid lines with errors
where one seem seems to account or explain another theme.
57. 7.1.4. Grounded theory
• Grounded theory offers a neat encapsulation of the
essence of interpretive enquiry: elect the ideas (the
theory) emerge from your immersion in the
situation rather than going in with fixed ideas (fixed
theory) but what is happening.
• Many of the assumptions behind grounded theory
seemed inappropriate and outdated now. In essence
the constant comparison method is a kernel of
grounded theory. The nuts and bolts of grounded
theory procedures complex and you're advised to
avoid them where possible.
• Stick to constant comparison method.
58. 7.1.5. Discourse and content
analysis
• Discourse analysis is the study of language in social
use. However it is spoken about in different ways in
different branches of social sciences, resulting in
confusion of the method.
• Psychologists think of discourse as the language that
goes on between people, tending to focus on small
units of language such as the choice of individual
words and micro analysis involved.
• Whereas sociologists tend to think of discourse as
forms of language that define social relationships
particularly power relationships between, and
among, people and look at macro analysis involved.
59. 7.1.5. Discourse and content analysis
• The term content analysis is sometimes used when the
analysis refers to the written text rather than the spoken
word.
• For simplicity's sake the general method in analysing an
interview is broadly the same as in the constant comparative
method.
• The difference is in the focus of the discourse analyst: rather
than being at the first level on the general ideas, the focus
tends to be on the use of particular words, phrases,
metaphors et cetera.
• In each case the discourse analyst will look to see how
notions constructed by the choice of words and language
form used in the discourse.
• Discourse analysis stresses the coding aspect of the
analysis of an interview, paying more attention to the choice
and use of words and phrases rather than the overall theme.
60. 7.2. ANALYSING NUMBERS
1. Statistics that describe
2. Statistics that understand the relationship between 2 variables
3. Statistics that help you to deduce or infer
61. 7.2. Analysing numbers
• Do not be intimidated by numbers! Statistics are not as hard
as you think; can be quite useful and help you determine
which research has found out.
• You need to understand that you know more about statistics
than you realise.
• Remind you the use of numbers in statistics comes in three
forms:
• 1. The categories of things: male or female. These are called
nominal or categorical data
• 2. Things we can put in order: first, second, third or top,
middle, bottom. These are called ordinal data. Although
there is an order indicated here there is no value implied
beyond this. Likert scale is an example of ordinal data
• 3. The everyday numbers: interval data because the
intervals between the numbers always the same
62. 7.2. Analysing numbers
• By understanding the difference between the
three levels of numbers you are able to utilise
these in a meaningful way without confusing
the data.
• Example you cannot multiply nominal data
• With small amounts of data you do not need to
make use of SPSS.
• Your Microsoft Excel spreadsheet is more than
sophisticated enough to handle the data and
any statistical analysis needed.
63. 7.2.1. Statistics that describe
• Descriptive statistics or about this implication, organisation,
summary and graphical plotting of numerical data.
• They are easy!
• Descriptive data covers questions such as "how many; how
often; how frequent"
• They are the simple statistics such as percentages and
averages
• Do try and make numbers in your dissertation meaningful to
the reader by using statistics at the most basic level.
The most common statistical data useful for all dissertations the
following:
Mean Mode Medium
Frequency distribution
Standard deviation
64. 7.2.2. Statistics that understand the
relationship between 2 variables
You may want to look at two features of the situation
and see whether the two are interrelated
You're exploring the concept of co-variance: how things
vary together.
In a silly example we can show that shoe size and
reading age co-vary: one goes up with the other !
You're attempting to describe the extent of the
connection between one variable and another: the
correlation coefficient. This will be a number between -
1 and +1. The nearer +1 the result is, the closer is the
relationship between the two sets of scores.
65. 7.2.3. Statistics that help you to
deduce or infer
Statistics that help you deduce or infer are called inferential statistics.
The former large part of statistics in social sciences.
They used particularly when we are trying to interpret the results of an
experiment.
What the tests do is to enable you to say whether the results you have
obtained are extendable beyond the date you have gathered in your
sample.
It answers the question: is the difference you have noted between the
experimental group and the control group one in which you can rely
for this purpose of extension, or is it one that may have occurred by
chance in your study? This involves the discussion of of
probability....the probability is less than 5 in 100 would have been by
chance! "p < 0.05"
This is all about significance testing: the figures relating to chance.
The two most frequently used in statistics are chi-squared and the t-
test.