PSYC 2230 Lecture Notes - Dependent And Independent Variables, Inter-Rater Reliability, Quantitative Psychology

22 views7 pages
Published on 26 Jun 2012
York University
PSYC 2230
of 7
Chapter 2: The Scientific Method in Psychology
Goal of quantitative/ psychology is the explanation of behaviour
Need language that is precise enough to be understood by others and
general enough to apply it to a variety of situations – this language takes the
form of explanations which are general statements about events that cause
phenomena to occur
Scientific method – agreed-upon approach to discovery and explanation
oSet of rules that dictate general procedures a scientist must follow in
oRules are based on logic and common sense
Three types of scientific research:
Naturalistic observations – observation of behaviour of people or animals in
their natural environment
oLeast formal and constrained by fewest rules
oProvide foundation of biological and social sciences
Correlational studies examination of relations between 2+ measurements
of behaviour or other characteristics of people or animals
oInvolve more formal measurements of environmental events, of
individuals’ physical and social characteristics, and of their behaviour
Experiments – study which researcher changes value of an independent
variable and observes whether this manipulation effects value of dependent
oOnly research which can confirm existence of cause-and-effect
relations among variables
Five steps summarizing the rules of scientific method:
1. Identify the problem and formulate hypothetical cause-and-effect relations
among variables
Involves identifying variables (behaviours and environmental and
physiological events) and describing relations among them in general
2. Design the experiment
experiments involve manipulation of independent variables and
observation of dependent variables
each variable must be operationally defined and independent variable
must be controlled so that it is only thing responsible for changes to
dependent variable
3. Perform experiment
organize material, train people to perform it, recruit volunteers,
randomly assign volunteers to experimental group of control group
4. Evaluate hypothesis by examining data from study
Involves special mathematical procedures used to determine whether
observed effect is statistically significant
5. Communicate results
Article, conferences, conventions
Replication – repetition of an experiment or observational study to see
whether previous results will be obtains; uncovers statistical anomalies
and incompetently conducted research
Great scientific research occurs as result of long-term research programs
in which findings are part of collective endeavour
Psychological research in Canada is supported by:
Natural Sciences and Engineering Research Council
Social Sciences and Humanities Research Council
Canadian Institutes of Health Research (Medical Research Council)
Step 1: Identifying the Problem
Hypotheses - statement, usually designed to be tested by an experiment,
that tentatively expresses a cause-and-effect relationship between 2+
Theory – set of statements designed to explain a set of phenomena,
proposes relationships among variables, and makes new predictions;
more encompassing than a hypothesis
way of organizing system of related hypotheses to explain some larger
aspect of nature
generates testable hypotheses - ones that can be supported or proved
wrong by scientific research
framework for most psychological research is larger in scope than
hypothesis but smaller in scope than full-fledged theory
Naturalistic Observations
Naturalistics – people who carefully observe animals in their own
environment, disturbing them as little as possible; observer remains in
Step 2: Designing an Experiment
Variables – anything capable of assuming any of several values (very in
Manipulate – setting values of an independent variable in an experiment
to see whether value of another variable is affected
Experimental Group – group of participants in an experiment, members
of which are exposed to particular value of independent variable, which
has been manipulated by researcher
Control Group – comparison group used in experiment, members of
which are exposed to naturally occurring or zero value of independent
Variable – anything capable of assuming any of several values
Independent variable – variable that is manipulated in experiment as
means of determining cause-and-effect
Dependent variable – variable that is measured in experiment as means
of determining cause-and-effect relations
value of dependent variable depends on value of independent variable
independent and dependent variables are categories into which
various behaviours are classified
experiment is performed by manipulating value of independent variable
and observing whether change affects dependent variable
Nominal Fallacy – false belief that one has explained causes of
phenomenon by identifying and naming it; (explaining lazy behaviour by
attributing it to “laziness”)
Operational Definitions – translation of generalities into specific terms;
definition of a variable in terms of operations the researcher performs to
measure or manipulate it
Validity -degree to which operational definition of a variable reflects
variable it is designed to measure or manipulate; how appropriate it is for
testing researcher's hypothesis
used to develop manipulations and measures of a vast array of
psychological concepts
if different investigators define variables in slightly different ways but
explanations yield similar results, we become more confident that we
are approaching good explanation of phenomena we are studying
confound – fail to distinguish
Confounding Variables – inadvertent simultaneous manipulation of more
than one variable; results of an experiment involving confounded variables
permit no valid conclusions about cause-and-effect
Counterbalancing – systematic variation of conditions in an experiment,
such as order of presentation of stimulus, so that different participants
encounter them in different orders; prevents confounding of independent
variables with time-dependent processes such as habituation or fatigue
Step 3: Performing an experiment
Reliability – repeatability of a measurement; likelihood that if
measurement was made again it would yield same value
Interrater Reliability – degree to which 2+ independent observers
agree in their ratings of another organism's behaviour
Random Assignment – procedure in which each participant has an
equally likely chance of being assigned to any of the conditions or groups of
an experiment
Used to avoid confounding participant characteristics with
manipulated values of an independent variable
Single-blind study – one participant receives a placebo – participant
doesn’t know value of independent variable
Double-blind study – neither participant and researcher know value
of independent variable
Correlational Studies