Philosophy of science 1 intro i and quantitative research
1. PHILOSOPHY INTRODUCTION
OF SCIENCE 1
Joseph Wright of Derby:
An Experiment With a Bird
in an Air Pump (1768)
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
2. PHILOSOPHY OF SCIENCE 1 INTRODUCTION
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
The Search Keywords: What is science, paradigm shifts, natural and social science,
for the Truth humanities, validation, falsification, induction, deduction
Introduction Scope: research;
qualitative and quantitative. Basic theory of research
Note on Literature
International class: Danish class:
This lesson is generally(!) based based on Lisa This lesson is generally(!) based based on
Bartolotti (2010): An Introduction to the Carsten Rønn (2006): Almen Videnskabsteori
Philosophy of Science. Polity, UK. for professionsuddannelserne. Alinea, DK.
Author’s profile: www.lisabortolotti.com Se også: www.samfundsviden.dk under
Philosophy ressource: http://www.iep.utm.edu/ teorier og metoder > videnskabsteori
6. PHILOSOPHY OF SCIENCE 1 INTRODUCTION
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
A Proper
Judgement Ancient / medieval times Renaissance and our time
Natural Science The discovery of the real The systematic obseervation of
and Social nature of things through reality in order to find regularity
Sciences philosophy and observation and cause‐effect of nature.
Method Classification of the order of Defining the quantitative traits
nature by looking at the of nature by meassuring,
qualitative aspects of things counting etc.
Truth Evidence Correspondence
10. PHILOSOPHY OF SCIENCE 1 INTRODUCTION
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
Validation
THE HP‐METHOD
Hypothetic‐ Hypotheses are possible causes, explanation of observations.
deductive
Reasoning The hypothesis tested first should be the most reasonable one.
Multiple hypotheses should be proposed whenever possible.
Hypotheses should be phrased such that predictions can be made.
Hypotheses can be ruled out with absolute certainty but cannot be proven
with absolute certainty.
12. PHILOSOPHY OF SCIENCE 1 INTRODUCTION
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
Validation
THE HP‐METHOD: Natural Sciences/positivism/verification
Hypothetic‐ Case: Semmelweis:
deductive http://www.indiana.edu/~koertge/X200Semm.html
Reasoning DK:
http://m.lilleor.dk/upl/9608/VidenskabslogikogSemmelveismoli1998pdf.pdf
Premis 1 (hypothesis): If A is a result of B, then the A stops when B is solved
by C (handwashing in the case of Semmelweis’ experiment).
Premis 2 (aiding hypothesis): We make sure C is conducted.
Premis 3 (observation): When C is conducted, B is dropping.
Conclusion: The difference between A and B is caused by … and can be
solved by C.
Critical view Not completely verified and true, but likely. It is easier to use this
experiment as a falsifiable answer (Karl Popper).
13. PHILOSOPHY OF SCIENCE 1 INTRODUCTION
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
Falsification
KARL POPPER
Falsifiable
Hypotheses Popper describes science as risky. We can only
temporarily prove something by letting some
premises become falsified:
I saw a white swan: All swans are white
Ok. Now I see a black swan: Swans can be black and
white. Hey! Now they are grey! (Etc.)
A single experience can contradict the prediction
based on a general hypothesis. This proves the
hypothesis as false. Popper claims, that only scientific
theories are falsifiable this way. It is not possible in,
say, psychoanalysis (says Popper)
14. PHILOSOPHY OF SCIENCE 1 INTRODUCTION
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
Exploring
Social Sciences Social facts are more complex. Often it’s a question of understanding the
phenomena rather than solving why it is occurring. It is almost impossible
to predict all social and cultural actions. Let alone the problem of
objectivity stands in the way for a “truth”. Life is more than the sum of all
its actions (holism). To understand these things we need to research
qualitative data. We must use our imagination and our empathy.
Critical view See Bartolotti (2010): p. 18–20.
15. PHILOSOPHY OF SCIENCE 1 INTRODUCTION
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
Science Cultures
Natural Science Social Science Humanities
Natural Hypothetic‐deductive method Hermeneutics
and Social Quantitative methods Qualitative methods
Sciences
Explanations Explaining causes Intentional
of causes by actions explanations
Makes laws Laws based on Formulates discourses
of universal validity statistical possibility with individual validity
Truth as correspondence Truth as meaning
and coherence
Critical view It’s almost never confined to one or the other. There can be a mix.
16. PHILOSOPHY OF SCIENCE 1 INTRODUCTION
INTRODUKTION TIL VIDENSKABSTEORI > BEGREBER > KVALITATIV OG KVANTITATIV RESEARCH
Science Cultures
Social Science Humanities
Area Nature Mankind, life
Natural of phenomena Reality “out there” Culture, philosophy, art
and Social
Sciences Focus The eternal, the non The historical view, the
changing, universal individual changes, unique
The objective, abstract and subjective concrete area
area (laws of nature)
Method Emperical observation Interpretation (hermenautics),
Classification
Describing Exploring
Quantitative (what) Qualitative (why)
Goal Explanation Understanding
Criteria of truth Truth as correspondence Truth as meaning/coherence.
18. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
The Research There are many ways to conduct a research. You often have to adjust your
Process objectives to the field you’re working in and to the environments and
Introduction people you are working with. The field you’re working in could be “website
usability” and the specific environments and persons (informants) you are
working with could be an office environment and its staff.
This is the first part of two lessons. During these lessons you will work
theoretically and practically with two main types of research processes:
the quantitative research process and the qualitative research process
Note on Literature:
This part of the lesson—on research—is based on Klaus Bruhn Jensen’s (ed.) book
A Handboook of Media and Communication Research: “The Quantitative Research Process”
by Barrie Gunter (chapter 13, p. 209–234). Routledge 2005.
On Google Books: http://books.google.com/books?id=Bt6kuYR‐mBsC&printsec=frontcover&source=gbs_ge_summary_r&cad=0#v=onepage&q&f=false
19. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Primary
Research Basic research can be divided into
desk research and field research.
Field Research Field research refers to the collection of new data through primary research.
Primary data That means direct contact with people through interviews, focus groups and
serveys.
Bigger and complex surveys are often done by bying this expertise from
companies specialized in conducting effective, reliable surveys.
20. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Secondary
Research
Desk Research Desk research relies on existing data and information, published on the net,
Secondary data in printed magazines or any other valid source (!)
Reading economy articles, trend news, articles based on web surveys and
reports from cultural and national organisations gives you a good picture of
your target group.
21. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative In this lesson we will focus on basic, primary research
Research
Introduction Remember: qualitative and quantitative approaches
to the Basic can both be primary and secondary:
Concepts
PR I M A RY DATA S ECO N DA RY DATA
Qualitative • Observations • Documents
Data • Interviews (open questions) • Notes (from secondary source)
• Movie recording (actively) • Letters
• Think aloud test, etc. • Sound and movie recordings
(others material)
• Artifacts
Quantitative • Closed questions • Articles and pictures, etc.
Data • Surveys
• Clearly defined objectives
in observations, etc. • Statistics
• Registrations, etc.
Artifacts, letters etc. can become quantifiable data for some purposes,
and statistics can become proof of qualitative arguments
22. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Research 1. The hypothesis … is a proposition to be tested. Hypotheses makes prognostications about the links
between variables. They propose, that under a set of conditions, if an independent variable
Basic Theory is manipulated in a certain way, or is assumed to have a certain strength, it may be
expected to exert a measurable impact on a designated dependent variable. Then it can be
tested, if this hypothesis can be proved or disproved. (Jensen 2005: 212).
2. The Variable … is gender, age, nationality etc.
… is the empirical representation of a …
Before the research
3. Independent Variables Concept and a… Construct (combination of concepts)
… can be manipulated by the A concept represents A combination of concepts used to
researcher. an abstract idea that define the characteristics of the
… means the way the research embodies the nature of individual users as grouped in
process is constructed to observable phenomena, concepts (in the example).
measure a response. or an interpretation of Heavy users may be described by such
why such phenomea concepts as sociability, tolerance for
4. Dependent Variables occur. strong stimulation, risk‐taking etc.
After the research
This is also what we call the
… is the measure of the Example: Individuals
constituent attributes, i.e. values or
outcome of the research are defined in terms of
categories into which variables can be
based on its construct and its their use of a media.
divided.
dependent variables, i.e. it is You define the notions
the information created by the of a “light user” and a
variables and the hypothesis. “heavy user” based on
media types, levels of
defined usage etc.
23. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative The Quantitative approach to research is based on the scientific tradition
Research of studying aspects of human reality with empirical proof.
Introduction
to the Basic Research in anthropology, economics, geography, linguistics, history, polital
Concepts science etc. can often be done by initially measuring hard facts.
Quantitative research often aims at closed questions—questions your
respondents can answer yes or no to; questions that are very narrow defined.
24. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Examples
Research
Introduction Numbers: You can measure how many women there are in a room
to the Basic and you can measure how many men there are.
Concepts
Specific actions: You can measure people’s actions (but not why they do it!)
Opinions: You can measure people’s opinions by asking closed questions:
“Do you think our prime minister will win the next election?”
This question would be followed up by additional (anonymous) information
from the informant: gender, age, city etc.
You can use these data as valuable information in your research objective.
25. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Variables: Concepts and Constructs
Research
Introduction The basic part of the quantitative research process is the notion of the
to the Basic variable. Variables are in this context the emperical representation of a
Concepts concept or a construct (→). Let’s first define concepts:
Concepts
A concept represents an abstract idea that embodies the nature of
observable phenomenon, or an interpretation of why such phenomea occur.
For example, individuals may be differentiated in terms of their use of
media: What you define as a “active user” could be distinguished from what
you define as a “passive user” (in questions based on how often a specific
media or an application is used).
Media usage can be linked to explain different behavioual patterns:
Media usage becomes an explanatory concept (Jensen 2005: 210).
26. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Single‐Concept Example (media usage as behavioural patterns)
Research
Introduction The observable phenomenon:
to the Basic More and more people over the age of 65 use FaceBook (FB).
Concepts source/desk research:
http://www.insidefacebook.com/2009/12/11/facebook‐a‐top‐destination‐for‐users‐over‐65/
The concept I want to use:
What is it to be “user”—it’s a bit too vague a concept! Just because you have a
FB account you’re not always active. I find it interesting to investigate the
single concept of the “active user” in the age group 65+.
First, I define the concept of a “heavy user” by describing precisely what this
implies in relation to skills in the use of FB. Second, I suggest the number of
visits on FB per day in order for one to be a “active user” of FB.
objective: How many “active users” in the age group 65+ are there on FB in
Denmark alone. I base this on my concept, my hypothesis (later in this slide)
and on a sampling of informants (say, 1000 FB‐users in the age group 65+)
27. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Constructs (combination/series of concepts)
Research
Introduction A construct comprises a combination of concepts.
to the Basic
Concepts This term can be used as a way of defining characteristics or actions
of individuals that are associated with their personality type.
For example, one personality type is defined as sensation‐seeking.
High‐sensation seekers generally need higher levels of environmental
stimulation than low‐sensation seekers.
High‐sensation seekers may be described by a series of such concepts as:
sociability, tolerance for strong stimulation, risk‐taking etc.
Constructs have a dimensional quality, so that individuals may be
classified (in this example) as high or low on the personality dimension
of sensation‐seeking (Jensen 2005: 210).
28. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Now, let us return to the notion of the variable:
Research
Introduction Variables
to the Basic A variable is an emperical representation of a concept or a construct.
Concepts
Variables provide operational measures that can be
quantified and manipulated by researchers.
The concepts of gender, age, economics, and personal behaviour are variables
that you can measure. These variables must be described in further detail as
concepts and/or constructs, as it’s demonstrated in the previous slides.
In other words: The variables get more substantial when they are applied
to a concept or a construct. Gender is interesting because of the construct
of the differences between certain actions of men and women.
29. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Variables and attributes
Research
Introduction Variables may be further defined and differentiated in terms of their
to the Basic constituent attributes.
Concepts
Attributes are values or categories into which variables can be divided:
In the case of the gender, there are two categories, male and female.
In the case of age, individuals can be differentiated by age group or actual age.
30. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Independent and dependent variables
Research
Introduction Variables can be further defined in terms of their relationship with each other.
to the Basic
Concepts Before the research: The independent variable (concept/construct/attributes)
can be manipulated by the researcher—it is meant to produce some measurable
response or outcome.
You design your variables (concept/construct/attributes) to fit your objective, for
example light user, active user etc.
After the research: The dependent variable is the measure of the response or
outcome. It is the obtained data which is treated as information.
BREAK
31. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Using a Hypothesis
Research
Introduction Quantitive research is primarily concerned with demonstrating cause‐effect
to the Basic relationships, and any research project starts by making a hypothesis.
Concepts
A hypothesis is a question―a proposition to be tested.
Hypotheses makes prognostications about the links between variables.
“(A hypothesis) propose, that under a set of conditions, if an independent
variable (…) is assumed to have a certain strength, it may be expected to exert
a measurable impact on a designated dependent variable.” (Jensen 2005: 212).
What does this mean?: The independent variable is the concept /construct
before the research―the part that has an objective and thus it can be
manipulated. There is of course a certain expectation to the outcome of this.
Because of this, the linkage to the data you obtain (dependent variable)
should make your objective stronger. Does it confirm your hypothesis?
32. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Reliability and validity
Research
Introduction Another important aim of a research is to determine its reliability.
to the Basic
Concepts To meet the demands for reliability and validity, you can …
repeate evidence/tests to show similar results over time
or in another context.
differentiate between the respondent (age, gender, education, etc.)
explain accurately how and why you conduct your research.
use good internal validity: the design of the research process must be free
from theoretical and methodological errors. Use validated theory.
The reliability concerns the dependability and consistency of the relationship
in one or between more variables. The validity indicates whether a measure
properly captures the meaning of the concept or construct it represents.
(Jensen 2005: 212).
33. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative The Quantitative survey―field Research
Research
The Survey Surveys collect data and facts using independent variables. After the survey,
make judgements based on the dependent variables.
A descriptive survey simply attempts to document current conditions.
Public opinion polls, for example, can rovide information about people’s
present attitudes on a specified topic. The concept of “fear of terrorism” can
be constructed through a variety of constructs from “no fear” to “very afraid,”
and the survey can gain answers from different age groups. The outcome is
descriptive and can of course be used ad an initial research narrowing down a
target group to be researched further on.
Analytical surveys also collect descriptive data, but attempt to go on to
examine relationships among variables in order to test research hypotheses.
A survey may assess the impact of an advertising campaign on public
awareness of a brand and changes in the market’s share of a product. Such
explantory surveys can also research social effects in‐ and because of media.
34. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Quantitative survey methods―field Research
Research
The Survey Telephone interviews:
Methods You can accomplish the data very quickly and directly.
It is cheap to conduct
The respondents can be reached globally.
Face‐to‐face interviews:
Short interviews as well as longer interviews can be conducted,
i.e. in a shopping mall or in a home.
Visual artefacts can be used.
Video‐ and audio techniques can be used.
Better personal credibility can be achieved by personal interview.
Web‐/mail‐/online‐ or paper questionnaire:
Simple forms can be filled out by respondents. By using web or mail the
statistical answers will be easy to monitor quickly afterwards.
35. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Sampling
Research
The Survey A key aspect of quantitative research is sampling.
Sampling
Samples may be constructed on either a probablity or non‐probablity basis:
A probability sample is selected according to mathematical guidelines,
i.e. each unit is known. Hereby the researcher can also calculate the amount of
sampling error.
Non‐probability sampling is often used in media research. People are
selected for study on the grounds that they are available, convenient to access
and prepared to participate.
Let’s look at aspects of these methods of sampling (next slides …)
36. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Sampling
Research
The Survey Non‐probability sampling can be related to …
A purposive sample. Often used in advertising research, this sampling
method selects respondents according to specific criterion, such as their
purchase of a specific product, when, where etc.
A quota sample. You match the percentage of male and female population in
relation to preferences. You can’t ask all the men and women in a country, so
you conduct your survey based on a representative quota, for example 300
men and 400 women (if there are more women matching this number in
percentage).
37. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Sampling
Research
The Survey Probability sampling can be related to …
Random sampling. The basic for of random sampling: Every individual or unit
in a population has an equal chance to be selected for a questionnaire.
Systematic random sampling. A criteria is fixed. For example, every tenth
person is selected from the population.
Stratified random sampling. More matching criterias are built into the
selection of the respondents. For example gender, age, ethnecity etc. Hence,
if 51 per cent of a population is female, random selection of females to the
sample will cease once the target as been reached, namely 510 females of a
target sample of 1000.
38. STUDY SHEET
RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Quantitative Describing Data
Research
The Survey Links: See more on percentage calculation here …
http://www.math.com/everyone/calculators/calc_source/percent.htm
http://www.easycalculation.com/statistics/statistics.php
http://www.euromonitor.com/
Don’t panic!
If you use an online survey tool all the calculations are done for you.
But you’ll have to do the thinking!
40. RESEARCH PROCESS 1 THE QUANTITATIVE RESEARCH PROCESS & BASIC THEORY
Research
Preparing for a 1. Make a good, validated hypothesis about a media trend—for example a
valid research specific type of use of a media technology. You are welcome to start with
desk research, for example:
http://rossdawsonblog.com/weblog/archives/2010/08/key_trends_in_m.
html
2. Aim this survey at a relevant target group.
3. Let your hypothesis be related to answers you can measure in an
affirmative or non‐affirmatiwe way (quantitative, closed questions).
Before doing this, be sure that you have defined precisely what concepts
and/or constructs to use. You can describe the user categories.
4. What kind of sampling method do you need to use?
5. Define variables to be used.
6. Consider validity—can you falsify any premises? How?