Introduction to Research Methods and Statistics Lectures
Lecture 1: Basic Principles of Scientific Research
Psychology = a scientific discipline that studies behaviour and mental processes (emotions,
thoughts) using scientific methods
A Short History of Methodology
BUDDHA and ARISTOTLE - asked questions about the causes of behaviour,thoughts and
emotions
Wundt, James, Watson
James McKeen Cattell - methodology in educational curriculum
Science = based on research
Methods and statistics = how to do research properly
What is Science?
- A method of creating knowledge using 3 tools ...
1) Systematic (set up experiment, testing, analyse data) empiricism
2) Public verification
3) Solvable problems (scientific)
A scientists 2 jobs :
1) Discovering and documenting new phenomena, patterns, and relationships
2) Developing and evaluating explanations explanations of the phenomena
Strategies in (behavioral) research
- Descriptive = describing thoughts, behaviours, feelings, etc
- Correlational = associations between variables
- Experimental = cause and effect, characteristics : manipulation, random assignment,
and experimental control
- Quasi experimental = as experimental, but less control and/or no random assignment
(real world)
The Empirical Cycle
De Groot (1961)
- Observation
- Induction
- Deduction
- Testing
- Evaluation
,OBSERVATION
- An idea for a research question
- Freedom of design (Leary p.15-16)
- E.g. scientist makes a specific observation (people fighting), gets an idea and raises
a question (why do people fight) - general
INDUCTION
- Formulating a general theory based on observable facts
- “Leap of faith” - cannot be sure theory from observation is the only one / correct one
(can never observe ALL the people) (plausible theory)
- Theory = set of propositions that attempts to explain the relationships amongst a set
of concepts
- E.g specific observation (people fighting) -> general theory (stressful situations)
- Logical theory … too broad to test (vague)
DEDUCTION
- General theory to more specific question
- Reasoning from general theory to specific research question /hypothesis
- If theory is true, research question follows LOGICALLY from theory
- Deduction = logic
- Research hypothesis = prediction following from theory
- General theory -> research hypothesis (specific / clear)
1) Conceptual = what is meant by concept (abstract)
Found in a dictionary (vague)
2) Operational = how is a concept measured or induced in a particular study
(concrete)
Research / in depth
TESTING
- Test a hypothesis by conducting an actual experiment
- Carry out research / plan (making adjustments)
1) Collect data
2) Analyse data
3) Drawing conclusions
- Based on analysis about the specific research question / hypothesis
EVALUATION
- What does the result tell us about the theory?
- Confirmation or falsification?
- Adjust, expand, improve theory?
- Critical review of my own study?
- Research hypothesis -> general theory
Proof, Disproof, and Scientific Progress
- Proof (“true”) : logically impossible
- Disproof (“untrue”) : practically impossible
- General theory always based on leap of faith
- Can we prove theory with itself? No.
, - Collect evidence to support theory (more evidence, stronger theory)
- Methodological pluralism = variability of methods
- Number of confirmations
- Stringency of tests
Explaining Variability : Variance
Aims of Behavioral Research
- VARIABILITY = Describing, predicting and explaining differences in behaviour and
mental processes between people
- VARIABLE = something that can vary
Between people (heights, fear, motivation)
Between situations (work vs private)
Over the course of time (child vs adult)
VARIANCE = a measure of variability
= evaluates scores in references to standard (the mean)
= S^2
Total Variance
1) Compute the grand mean (GM)
2) Subtract GM from each observation (=deviation sore)
3) Square each deviation score
4) Add all squared deviations (SS(TOTAL))
5) Divide by n-1
, Add all scores and divide by how many there are
Score of individual minus grand mean
Square that number (otherwise result s always 0)
Sum of squared scores
Divide summed score by n-1 (n=number of scores)
Splitting up total variance
Total variance = all differences between observations (what we want to explain)
Systematic variance = differences between groups related to a certain variable (which we
can account for)
Error variance = unexplained differences (what we CANNOT account for)
SYSTEMATIC VARIANCE (between groups)
1) Compute the group mean (and the grand mean)
2) Subtract grand mean from each group mean
3) Square each deviation
4) Multiply by the number of observations in that group
5) Add all squared deviations
6) Divide by n-1
ERROR VARIANCE
1) Compute group means
2) Subtract group mean from each observation in that group
3) Square each deviation
4) Add all deviations (SS(WITHIN)
5) Divide by n-1
Total variance = systematic variance + error variance
Lecture 1: Basic Principles of Scientific Research
Psychology = a scientific discipline that studies behaviour and mental processes (emotions,
thoughts) using scientific methods
A Short History of Methodology
BUDDHA and ARISTOTLE - asked questions about the causes of behaviour,thoughts and
emotions
Wundt, James, Watson
James McKeen Cattell - methodology in educational curriculum
Science = based on research
Methods and statistics = how to do research properly
What is Science?
- A method of creating knowledge using 3 tools ...
1) Systematic (set up experiment, testing, analyse data) empiricism
2) Public verification
3) Solvable problems (scientific)
A scientists 2 jobs :
1) Discovering and documenting new phenomena, patterns, and relationships
2) Developing and evaluating explanations explanations of the phenomena
Strategies in (behavioral) research
- Descriptive = describing thoughts, behaviours, feelings, etc
- Correlational = associations between variables
- Experimental = cause and effect, characteristics : manipulation, random assignment,
and experimental control
- Quasi experimental = as experimental, but less control and/or no random assignment
(real world)
The Empirical Cycle
De Groot (1961)
- Observation
- Induction
- Deduction
- Testing
- Evaluation
,OBSERVATION
- An idea for a research question
- Freedom of design (Leary p.15-16)
- E.g. scientist makes a specific observation (people fighting), gets an idea and raises
a question (why do people fight) - general
INDUCTION
- Formulating a general theory based on observable facts
- “Leap of faith” - cannot be sure theory from observation is the only one / correct one
(can never observe ALL the people) (plausible theory)
- Theory = set of propositions that attempts to explain the relationships amongst a set
of concepts
- E.g specific observation (people fighting) -> general theory (stressful situations)
- Logical theory … too broad to test (vague)
DEDUCTION
- General theory to more specific question
- Reasoning from general theory to specific research question /hypothesis
- If theory is true, research question follows LOGICALLY from theory
- Deduction = logic
- Research hypothesis = prediction following from theory
- General theory -> research hypothesis (specific / clear)
1) Conceptual = what is meant by concept (abstract)
Found in a dictionary (vague)
2) Operational = how is a concept measured or induced in a particular study
(concrete)
Research / in depth
TESTING
- Test a hypothesis by conducting an actual experiment
- Carry out research / plan (making adjustments)
1) Collect data
2) Analyse data
3) Drawing conclusions
- Based on analysis about the specific research question / hypothesis
EVALUATION
- What does the result tell us about the theory?
- Confirmation or falsification?
- Adjust, expand, improve theory?
- Critical review of my own study?
- Research hypothesis -> general theory
Proof, Disproof, and Scientific Progress
- Proof (“true”) : logically impossible
- Disproof (“untrue”) : practically impossible
- General theory always based on leap of faith
- Can we prove theory with itself? No.
, - Collect evidence to support theory (more evidence, stronger theory)
- Methodological pluralism = variability of methods
- Number of confirmations
- Stringency of tests
Explaining Variability : Variance
Aims of Behavioral Research
- VARIABILITY = Describing, predicting and explaining differences in behaviour and
mental processes between people
- VARIABLE = something that can vary
Between people (heights, fear, motivation)
Between situations (work vs private)
Over the course of time (child vs adult)
VARIANCE = a measure of variability
= evaluates scores in references to standard (the mean)
= S^2
Total Variance
1) Compute the grand mean (GM)
2) Subtract GM from each observation (=deviation sore)
3) Square each deviation score
4) Add all squared deviations (SS(TOTAL))
5) Divide by n-1
, Add all scores and divide by how many there are
Score of individual minus grand mean
Square that number (otherwise result s always 0)
Sum of squared scores
Divide summed score by n-1 (n=number of scores)
Splitting up total variance
Total variance = all differences between observations (what we want to explain)
Systematic variance = differences between groups related to a certain variable (which we
can account for)
Error variance = unexplained differences (what we CANNOT account for)
SYSTEMATIC VARIANCE (between groups)
1) Compute the group mean (and the grand mean)
2) Subtract grand mean from each group mean
3) Square each deviation
4) Multiply by the number of observations in that group
5) Add all squared deviations
6) Divide by n-1
ERROR VARIANCE
1) Compute group means
2) Subtract group mean from each observation in that group
3) Square each deviation
4) Add all deviations (SS(WITHIN)
5) Divide by n-1
Total variance = systematic variance + error variance
