Statistics in Psychology (Chapter 1)
Variability in psychology: “it is not the empirical nature of psychology per se that gives rise to the
need for statistics and research methods; rather, it is the fact that psychology is an empirical discipline
in a world of variability, uncertainty and inaccuracy. Given the nature of living things, such variability
is inherent in all human, biological and social sciences.”
→ Psychology relies on statistics and research methods not just because it studies real-world
phenomena, but because it deals with variability, uncertainty, and inaccuracy. Since human
behaviour and social interactions are naturally inconsistent, variability is unavoidable in
psychology and other life sciences.
Research methods: refers to a set of procedures and techniques for data collection in situations of
variability
Statistics: a set of techniques for exploring, summarising and making sense of the data collected
1.1 Causes of Variability
4 broad clusters:
1. Measurement variation: errors in measurement
a. sensitivity/calibration of measuring instruments (ie measurement is only as good as
the instrument that produces it)
b. Inaccuracies in measuring instruments → systematic or unsystematic
2. Situational variation
a. Affected by environmental conditions (eg time of day, room temperature, appearance
of researcher, test instructions)
3. Individual variation → most important source of variability in psychological research
a. Physical variation (eg height, weight)
b. Psychological variation (eg big 5 personality, HEXACO)
c. Tendency for individuals to vary in their own behaviour or performance from one
occasion to another
4. Sample variation
a. Differences of individuals that make up the sample
b. Some samples reflect parent population more fairly than others → more
representative/generalisable
1.2 The Empirical Process
,1.3 Interrelationship of Theory and Experimentation
For research to move forward…
● Hypotheses must be rooted in current knowledge of the topic
● Data collected must allow predictions contained in hypotheses to be tested
● Outcomes must be assimilated into current body of knowledge
Hypothetico-deductive method: scientific inquiry that proceeds through several stages of the empirical
process, namely…
1. Formulation of a hypothesis based on existing theory
2. Data collection
3. Statistical analysis
4. Conclusion (refute existing theory through falsification?)
5. Modification of existing theory as appropriate
1.4 Observation vs Intervention
Experimental method: researcher undertake experiments, typically laboratory-based and involving
active intervention and control of experimental conditions, with relatively small numbers of
participants
correlational/observational method: researchers observe or measure people’s behaviour without
direct intervention or control of conditions, typically involving relatively large samples of
participants, and often in natural settings
, Variables (Chapter 2)
2.1 Ways of Measuring People (and Animals)
Note: this is a representative and not an exhaustive list of measurement possibilities
1. Direct physical measurements
○ eg height, weight, grip strength, heart rate, galvanic skin response, brain activity scan
2. Observation
○ Can be from naturalistic or highly controlled environments
○ Can be with or without direct intervention by experimenter
○ Can be qualitative (described in verbal terms) or quantitative
3. Performance measurements
○ Aspects of performance in a variety of tasks (eg speed, accuracy, productivity) can be
measured
○ Eg typing skills → measured by amount of text typed in conjunction with number of
errors made
4. Self-reports
○ Eg perception experiment: participant asked to judge whether 2 stimuli are the same
or different in some dimensions
○ Questionnaires
5. Subjective ratings
○ Ask a panel of people to make subjective judgements and combine into a single figure
2.2 Nature of Data Generated
Note: the way in which number is used defines the properties of the number, and determines which
arithmetical operations can validly be applied
Eg It makes sense to say that 4 bottles of milk is 1 fewer than 5 bottles of milk, and twice as
many as 2 bottles. It makes sense to say that 4°C is 1°C less than 5°C, but actually it does not
make sense to say that 4°C is twice as warm as 2°C.
Eg The player in a football team wearing the number 4 shirt is not twice as anything as the player
wearing the number 2 shirt
Nature of numbers
1. Natural numbers → counting discriminable entities (eg number of times something happen)
2. Other numbers represent quantities that can vary continuously and fill in gaps between natural
numbers (ie whole numbers)
, 2.3 Levels of Measurement
4 levels of measurement:
1. Nominal scale: numbers used merely as labels for categories to divide (eg nationality)
2. Ordinal scale: numbers have a definite order and appropriate inferences can be drawn from
that order (eg sports ranking)
3. Interval scale: implies all property of ordinal scale, and additional property that
differences/intervals are considered equal (eg temperature)
● Note: while we can assume equal interval, does not make sense to compare points on
the scale in terms of ratios
4. Ratio scale: implies all property of interval scale, and additional property of an absolute zero
(eg weight)
2.4 Variables
Variable: refers to any aspect of interest that varies among people, animals or other units of analysis
(eg countries)
Eg we can conceive of aggressiveness as something that a person, an animal or a country
possesses to a greater or a lesser degree
Note: variable definitions and measurement depend on cultural and scientific perspectives
● Once a concept is seen as a variable, people assume it has measurable properties
○ Eg tolerance is often viewed as a trait that people possess in different degrees, can be
measured, and compared numerically
Examples of Variables & Measurement Methods:
● Handedness: not simply left- or right-handed; assessed through tasks (eg Edinburgh
Handedness Inventory)
● Reaction time: measured in experiments where participants respond to stimuli as quickly as
possible
● Attitude toward computers: evaluated via questionnaires with a 1 - 5 rating scale, where
responses are scored based on positive or negative attitudes
● Intelligence: traditionally measured through IQ tests, though there is debate over cultural bias
and whether intelligence can be reduced to a single score
2.5 Variables under Experimental Control
Dependent variables: the variable that changes as a result of the independent variable manipulation
independent variables: a variable whose variation does not depend on that of another, and is often
manipulated by the researcher
Note: blurring of the two types of variables can occur
Eg when the researcher uses a measure of performance to categorise participants into one of a
number of groups defining an independent variable → for any given person, the value of the
independent variable is inherent to that person, and not under the control of the experimenter
→ dependent variable used to create the independent variable
Variability in psychology: “it is not the empirical nature of psychology per se that gives rise to the
need for statistics and research methods; rather, it is the fact that psychology is an empirical discipline
in a world of variability, uncertainty and inaccuracy. Given the nature of living things, such variability
is inherent in all human, biological and social sciences.”
→ Psychology relies on statistics and research methods not just because it studies real-world
phenomena, but because it deals with variability, uncertainty, and inaccuracy. Since human
behaviour and social interactions are naturally inconsistent, variability is unavoidable in
psychology and other life sciences.
Research methods: refers to a set of procedures and techniques for data collection in situations of
variability
Statistics: a set of techniques for exploring, summarising and making sense of the data collected
1.1 Causes of Variability
4 broad clusters:
1. Measurement variation: errors in measurement
a. sensitivity/calibration of measuring instruments (ie measurement is only as good as
the instrument that produces it)
b. Inaccuracies in measuring instruments → systematic or unsystematic
2. Situational variation
a. Affected by environmental conditions (eg time of day, room temperature, appearance
of researcher, test instructions)
3. Individual variation → most important source of variability in psychological research
a. Physical variation (eg height, weight)
b. Psychological variation (eg big 5 personality, HEXACO)
c. Tendency for individuals to vary in their own behaviour or performance from one
occasion to another
4. Sample variation
a. Differences of individuals that make up the sample
b. Some samples reflect parent population more fairly than others → more
representative/generalisable
1.2 The Empirical Process
,1.3 Interrelationship of Theory and Experimentation
For research to move forward…
● Hypotheses must be rooted in current knowledge of the topic
● Data collected must allow predictions contained in hypotheses to be tested
● Outcomes must be assimilated into current body of knowledge
Hypothetico-deductive method: scientific inquiry that proceeds through several stages of the empirical
process, namely…
1. Formulation of a hypothesis based on existing theory
2. Data collection
3. Statistical analysis
4. Conclusion (refute existing theory through falsification?)
5. Modification of existing theory as appropriate
1.4 Observation vs Intervention
Experimental method: researcher undertake experiments, typically laboratory-based and involving
active intervention and control of experimental conditions, with relatively small numbers of
participants
correlational/observational method: researchers observe or measure people’s behaviour without
direct intervention or control of conditions, typically involving relatively large samples of
participants, and often in natural settings
, Variables (Chapter 2)
2.1 Ways of Measuring People (and Animals)
Note: this is a representative and not an exhaustive list of measurement possibilities
1. Direct physical measurements
○ eg height, weight, grip strength, heart rate, galvanic skin response, brain activity scan
2. Observation
○ Can be from naturalistic or highly controlled environments
○ Can be with or without direct intervention by experimenter
○ Can be qualitative (described in verbal terms) or quantitative
3. Performance measurements
○ Aspects of performance in a variety of tasks (eg speed, accuracy, productivity) can be
measured
○ Eg typing skills → measured by amount of text typed in conjunction with number of
errors made
4. Self-reports
○ Eg perception experiment: participant asked to judge whether 2 stimuli are the same
or different in some dimensions
○ Questionnaires
5. Subjective ratings
○ Ask a panel of people to make subjective judgements and combine into a single figure
2.2 Nature of Data Generated
Note: the way in which number is used defines the properties of the number, and determines which
arithmetical operations can validly be applied
Eg It makes sense to say that 4 bottles of milk is 1 fewer than 5 bottles of milk, and twice as
many as 2 bottles. It makes sense to say that 4°C is 1°C less than 5°C, but actually it does not
make sense to say that 4°C is twice as warm as 2°C.
Eg The player in a football team wearing the number 4 shirt is not twice as anything as the player
wearing the number 2 shirt
Nature of numbers
1. Natural numbers → counting discriminable entities (eg number of times something happen)
2. Other numbers represent quantities that can vary continuously and fill in gaps between natural
numbers (ie whole numbers)
, 2.3 Levels of Measurement
4 levels of measurement:
1. Nominal scale: numbers used merely as labels for categories to divide (eg nationality)
2. Ordinal scale: numbers have a definite order and appropriate inferences can be drawn from
that order (eg sports ranking)
3. Interval scale: implies all property of ordinal scale, and additional property that
differences/intervals are considered equal (eg temperature)
● Note: while we can assume equal interval, does not make sense to compare points on
the scale in terms of ratios
4. Ratio scale: implies all property of interval scale, and additional property of an absolute zero
(eg weight)
2.4 Variables
Variable: refers to any aspect of interest that varies among people, animals or other units of analysis
(eg countries)
Eg we can conceive of aggressiveness as something that a person, an animal or a country
possesses to a greater or a lesser degree
Note: variable definitions and measurement depend on cultural and scientific perspectives
● Once a concept is seen as a variable, people assume it has measurable properties
○ Eg tolerance is often viewed as a trait that people possess in different degrees, can be
measured, and compared numerically
Examples of Variables & Measurement Methods:
● Handedness: not simply left- or right-handed; assessed through tasks (eg Edinburgh
Handedness Inventory)
● Reaction time: measured in experiments where participants respond to stimuli as quickly as
possible
● Attitude toward computers: evaluated via questionnaires with a 1 - 5 rating scale, where
responses are scored based on positive or negative attitudes
● Intelligence: traditionally measured through IQ tests, though there is debate over cultural bias
and whether intelligence can be reduced to a single score
2.5 Variables under Experimental Control
Dependent variables: the variable that changes as a result of the independent variable manipulation
independent variables: a variable whose variation does not depend on that of another, and is often
manipulated by the researcher
Note: blurring of the two types of variables can occur
Eg when the researcher uses a measure of performance to categorise participants into one of a
number of groups defining an independent variable → for any given person, the value of the
independent variable is inherent to that person, and not under the control of the experimenter
→ dependent variable used to create the independent variable
