Bridging the Gap - Qualitative & Quantitative Methodologies Slides

Download Report

Transcript Bridging the Gap - Qualitative & Quantitative Methodologies Slides

Bridging the gap:
Qualitative & Quantitative methodologies
Jo Ferrie Interim Director of Graduate Training
School of Social & Political Sciences
Venues and moodle codes
• SSS1
– Lecture 12-3pm Mondays (lecture 12-2.15ish (mandatory), then a
revision session (optional))
– Macgregor Building on the Western Infirmary Site(Main lecture
theatre)
– Tutorials will be assigned to you at the first lecture
– Moodle code – sss1
• Qualitative Methods
– Lecture 5:30-6:30 Tuesdays
– Adam Smith Lecture Theatre Room 1115
– Tutorials begin the week after (week beginning: 30th Sept), groups
will be assigned to you at the first lecture
– Moodle code – qual2012
Overview
• Qualitative & Quantitative approaches to social science
research – general reflections
• Quantitative research methods
– Use of statistical data in the media
– Mixed methods
Why so important?
• Unsettling the way in which you go about ‘doing’ your
research;
• Challenging disciplinary conventions;
• Potential value of mixed method approaches
• Appreciation of other types of research /approaches
Search for truth …
“Each society has its regime of truth, its ‘general politics’ of
truth: that is, the types of discourse which it accepts and
makes function as true; the mechanisms and instances
which enable one to distinguish true and false statements,
the means by which each is sanctioned; the techniques
and procedures accorded value in the acquisition of truth;
the status of those who are charged with saying what
counts as true” (Foucault 1980, 131).
– Foucault M. 1980. Power/knowledge: Selected interviews and
other writings 1972–77, ed. C Gordon. New York: Pantheon.
Qualitative or Quantitative approaches to
research?
•
•
•
•
•
“Lies, damned lies, and statistics”
‘Physics envy’ – purported prestige of the ‘hard’ sciences
Rigour/reliability – qualitative/subjective (?) etc.
“Scientists are not and cannot be concerned with the individual case. They
seek laws, systematic relations, explanations of phenomena. And their results
are always statistical.” (Kerlinger 1979: 270)
“The degree to which the observations can be quantified (translated into
numbers) is often a good index of the maturity of a science” (Mussen, Conger
& Kagan, 1977: 13)
Understanding human nature vs. understanding the physical world(?)
•
Mixed methods – best of both worlds?
•
5 minute discussion
• Watch the excerpt from Moneyball
– Reflect on the different styles that are brought to play – subjective
judgement vs. statistics etc. and associated claims to ‘truth’
– Reflect on the way in which individuals react to different claims to
insight, ‘truth’ etc.
Moneyball
•
•
•
•
2 minute scene
http://www.youtube.com/watch?v=HiB9L3dG-Aw
The baseball team has limited funds
The team’s general manager (Billy Beane) is trying to
approach the issue of which players to buy from a
different angle – one based on statistics/numbercrunching.
• Challenges the usual way in which things have been done
i.e. judgement based on subjective factors, experience,
‘feel’ etc.
• http://www.youtube.com/watch?v=yGf6LNWY9AI
• With your neighbour discuss:
– Your preference for Qualitative and/or Quantitative research
methodologies
• Indeed, do you think about social science research
methodologies in this way?
– Relative strengths of Qualitative & Quantitative research
methodologies
Let’s dwell on statistics...
can we find truth in statistics?
•
•
•
•
•
Yes
But … we need to know how to read them
Drinking beer reduces risk of coronary heart disease
Eating oily fish reduces risk of stroke
Eating five pieces of fruit and vegetables increases your
health rating
– Really?
– How can we be sure who to trust?
– Which truth is the real truth?
Making the case for quantitative methods
• Ontological
– Assumption that the social world is in fact a mathematical problem that
can be solved to produce universal laws
• Epistemological
– Research data should be quantitative in order to afford comparisons
across studies and theories
• Technical
– Statistical techniques are power tools for handling large amounts of data
• Rhetorical (informing/influencing the audience)
– Hard quantified facts are more convincing to the public – more
trustworthy
Positivism
• Application of the methods of the natural sciences to the study of
social reality
– Only knowledge confirmed by the senses is knowledge
– Positivism aimed to provide the laws of society and the possibility
of socially engineering society
– Laws are assessed via hypothesis testing
– ‘Value free’ and ‘objective’ conduct
– “The real power of man to modify phenomena at will can only
result from a knowledge of their natural laws” (Comte, 1974: 142).
Comte, A. (1974) The Essential Comte: Selections from Course de
Philosophie Positive (S. Andreski, Ed.) New York: Barnes & Noble
Main ‘requirements’ of quantitative methods
1. Theory
2. Hypothesis
3. Data Collection
4. Findings
5. Hypotheses confirmed
or rejected
• Requires a greater degree of
objectivity compared with qualitative
methods
– Findings will be scrutinised
– Findings should be replicable
• Usually deductive as it relies on
relationship between theory and
research in order to develop a
hypothesis that is then proved or
disproved by data
Figure: The process of reduction, Bryman,
2008: 10
• How statistics are used daily?
Examples from the world of sport
• Descriptive statistics often used to
record sporting achievement
• Tables include
•
•
simple counts
basic calculations e.g. goal difference
• Replicated across all walks of life
•
•
exam league tables
unemployment counts
• Not scary!
Do statistics always tell the whole story? (thanks to Cat Nixon
MRC)
Using the London 2012 Olympic medal table
• Olympic medal table example of “count” variable
•
•
Records number of medals won by colour
Tells us that the USA was most successful at the Olympics
• But what if the medal table isn’t telling the whole story?
•
•
•
Is it fair to just “count” medals?
What if we compared how countries did based upon population size and GDP?
Do you think the USA would still have come first in the table?
• Let’s have a look shall we….
•
Olympic table normally ordered by medals won and then sorted by Gold Medals
•
For simplicity let’s order by total number of medals won
•
Can see that USA is most successful team
Country
No.
medals
won
Official
Rank
Population
size
GDP
($bn)
Medals per
100,000
population
Medals
per $1bn
GDP
USA
104
1
309,349,000
15094
0.034
0.007
China
88
2
1,338,300,000
7298
0.007
0.012
Jamaica
12
21
2,702,000
15
0.444
0.800
Grenada
1
58=
104,000
0.8
0.961
1.250
Bahamas
1
58=
64,600
6
1.548
0.170
Botswana
1
69
2,007,000
18
0.050
0.060
Medals per 100,000 population = (no. medals won/population size)100000
Medals per $1bn GDP = no. medals won/GDP
Logarithmic adjustment for GDP and weighted by medal colour
Logarithmic adjustment for population size and weighted by medal colour
So what does it mean?
• How data are recorded and analysed affect how results are interpreted
• “Counting” shows us that the USA were most successful at London 2012
• But when you consider other factors a different picture emerges:
•
•
•
adjusting for GDP means that the USA would have come 66th
adjusting for population size means that the USA would have come 47th
Grenada could be considered most successful country at Olympics!
• Why does this matter? As social scientists we try to explain differences in
behaviours, outcomes etc by looking at the differences between and within
groups, not just by counting the number of times an outcome occurs.
Do you always believe what you read?
Yet another example from the world of sport
• Statistics are often our main source of information on a topic
• Data are easily manipulated to form arguments
• Data are often misreported within the wider press
- Creates “knowledge-base” that becomes “fact-base”
• Having a basic understanding of both qualitative and quantitative methods
allows us to move beyond merely accepting data and ask questions about the
validity of evidence
What the newspapers say….
But what do the data say?
• 118 medals won by 113
athletes
All
State
Private
Total
118
75
43
Gold
43
30
13
Silver
29
17
12
Bronze
46
28
18
• If we know what schools the
athletes attended then we can
calculate %age of medals won
by each group
• Is there evidence to support the
claims made by the press?
Number of athletes who won a medal at Olympic
Games by education type
100
90
80
70
60
50
40
30
20
10
0
36
30
41
39
59
61
Silver
Bronze
70
64
Total
Gold
Is it really that simple?
• There appears to b e evidence to support the statement that 4 in 10
medallists educated at private schools
• BUT there are approx. 9,500,000 school children in the UK and only
6.5% (617500) of which attend private schools
• What if we scaled the results to reflect this by creating a “medals won
per 10000 children” figure?
• Do you think you would be more likely to win a medal at a private or
state school?
•
Medals won per 10000 children = (medals won/no. educated in 2012)10000
Medals
won
No. educated in
2012
Medals won per
10000 children
Private
43
617,500
0.7
State
75
8,882,500
0.08
•
Results show that the number of medals won per 10000 children were higher
for those in private education
•
Dividing 0.7 by 0.08 we can conclude that based upon population size, being
educated in a private school makes you nearly 9 times more likely to win a
medal!
A real world example of mixed methods
•
PhD study funded by Chief Scientist Office for Scotland
•
Aims of study are to:
–
–
–
•
explore why sexual health of young people looked after by the state is worse than that of general population
identify how looked after young people learn about sexual health and use services
establish what role carers can play in improving outcomes for looked after young people
Mixed methods approach
–
–
–
Secondary data analysis of nationally representative sexual health dataset
Systematic literature review
In-depth and semi-structured interviews with looked after young people, carers and social workers
•
Sexual health dataset and interviews mainly focused on sexual health, but data exists
on wider health risk behaviours
•
Going to present data on substance use
Who are looked after children? (Cat Nixon
again, MRC)
•
Young people aged 0-18 who are looked after by the state due to difficulties within the
family
•
Reasons for being looked after include:
–
–
–
–
–
–
•
physical, emotional and sexual abuse
parental ill-health
child requires specialist care for medical reasons
parental substance use
persistent truanting from school
involvement in youth justice system
Small population but associated problems more likely
–
–
–
–
1.4% of 0-18 year olds in Scotland are looked after each year by local authorities
significantly more likely to be excluded from school,
higher rate of mental health problems
care history associated with homelessness, early parenthood and involvement in sex-working
Prevalence of health risk behaviours in children looked
after by state and those living with at least one biological
parent
100
90
80
70
60
50
40
30
20
10
0
29
18
21
63
43
50
Non-LAC
LAC
•
Looked after young people are significantly more likely to:
–
–
–
•
get drunk at least once a week
smoke cigarettes
smoke cannabis once a week
Literature suggests that the adoption of these behaviours is associated with:
–
–
–
–
deprivation
low levels of parental supervision
disengagement from education
having ‘risky’ peers
•
All of which are seen in higher rates amongst looked after young people
•
Adjusting for the effects of these in statistical models explains difference in behaviour
between looked after young people and rest of population
•
But what it doesn’t tell us is why they use substances at a higher rate
Qualitative interviews
• Interviews conducted with 33 young people aged 14-23
between August and December 2011
• Combination of in-depth and semi-structured interviewing
used to accommodate differences in educational,
emotional and communication differences
• Thematic analysis
Reasons for drug and alcohol use
• Young people often exposed to drugs and alcohol by
other young people they were living with
• Drugs and alcohol used to:
–
–
–
–
protect against “bullying” from other young people
“fit in” with other young people
to escape from the past
to block out being a looked after child
at first it was cannabis, poppers, aerosols, all the kind of stupit stuff
but then I started going out with a few of the people when I moved
into (the unit) and you started going out at the weekends and there
was eccies flying aboot and there was cocaine flying aboot and
speed, I mean I was only 12 year old and I was taking them all just to
show I wasnae a daft wee lassie… I was just trying to act the big hard
brass Andrea which I was at one point but that’s what everyone
thought but I was a vulnerable wee lassie on the inside who just
wanted to go out and drink and dae drugs tae just try to block it all
oot”.
Andrea, aged 16
“I’d just take anything, anything we could get
our hands on. We’d smoke it all and snort it all or
whatever and you used to feel really good, but then
the next day you’d just think ‘why am I doing this?’,
but it was really, I just think it was getting away
from everything that made it worthwhile.”
Sophie aged 14