Transcript Psychology 598 Research Methods

Psychology 598
Research Methods
Carolyn R. Fallahi, Ph. D.
Class #1 and 2
1
Case Study
 The study of one subject.
 Discuss Freud example.
2
Single-Case and Small-N research
designs
 Hermann Ebbinghaus example.
 Single-case designs.
3
The Single case design
 A-B-A design


A phase = baseline or pretreatment period
B phase = the introduction of the independent
variable.
4
Single case designs
 A-B design, the DV measured repeatedly
throughout the pretreatment and treatment
phases of the study.
 In the A-B-A design, the treatment is
withdrawn at the end, and the behavior is
measured.
5
Single case designs

A-B-BC-B design, B and C refer to 2
therapeutic conditions.
6
Single Case Designs
 A-B-A-B design.

2 occasions (B to A and then A to B) for
demonstrating the positive effects for the
treatment variable.
7
Correlational versus Causal Research



Correlational Research: associational
research
Not causation
Causal – comparative research
8
Survey Research
 The enumerative survey

Purpose: to count (enumerate) a
representative sample and then make
inferences about the frequencies of
occurrence in the population as a whole.
9
The Interview
 involves having the researcher ask questions
directly of the subjects.
10
Interview
 Open-ended versus structured interviews



Open-ended participants can expand on their
answers, to express feelings, motives, or
behavior quite spontaneously.
“Tell me in your own words how you felt when
________?”
critical incident technique.
11
Interview
 “Think about the last time you saw your mother drink
herself under the table. What thoughts were
associated with that incident?”
 Tell me exactly what you did to deal with this
situation?
 In contrast to open-ended questions, structure
(closed) questions are those with a clear cut
response option. For example... There are often
many reasons why a child does not do well in school.
Motivation can sometimes be an issue. Of the
following statements, please indicate which applies to
your child.
12
Interview
 1. tries very very hard
 2. Tries somewhat more than the average
student
 3. tries about like the average student
 4. tries somewhat less than the average
student
 5. doesn’t try at all.
13
Qualitative research
 Qualitative research: using words to
describe. For example, how do counselors
specializing in substance abuse deal with a
defensive patient?
14
Quantitative Research / Ethnographic
study
 Quantitative Research: How well, how much,
numbers.
 Ethnographic study
 Ethnographic record
15
Historical Research
 Historical research


Biography
Phenomenology – research that focuses on
particular issue.
16
Historical research / Action Research
 We are trying to look at a historical incident
and see the effects on people and variables.
 Action Research: change conditions within a
particular situation. We are not caring about
generalizing to other situations.
17
Meta-analysis
 Look at all studies within a topic and
statistically analyze the results across the
studies.
 Example: Rosenthal study
18
Conducting Research
 Problem Statement
 Hypothesis

Operational definition versus constitutive
definition



Constitutive definition – is basically a dictionary
definition.
You could clarify the hypothesis by example.
Operational definition
19
Conducting Research
 Variables
 Literature Review
 Sample/population
 Instruments
 Procedure
 Results
 Conclusion/Discussion
20
Variables
 Variable: a noun that stands for a variation
within a class of objects




For example: gender, motivation
Example: reinforcement: 3 types: verbal
praise, money, points on examination.
Quantitative Variables: On a continuum,
example weight and height.
Categorical variables: DK vary in degree or
amount, but are qualitatively different. For
example, gender, religious preference.
21
Variables
 IV




A True study focuses on the question: “How
things are and how they got to be that way?”
DV: refers to the status of the effect of
outcome in which the research is interested.
Example: Jogging makes you feel better.
The IV would be jogging status (jogging or not
jogging) and the DV would be feeling status
(feeling better or not feeling better).
22
Variables
 Extraneous variables or control variables

IVs that have not been controlled
23
Variables
 Another example: new medication …
Schizerall … design the study with them and
talk about potential extraneous variables.
24
Hypotheses
 Two types: operational and theoretical
25
The Null hypothesis in significance
testing
 The lingo used in an experiment



Ho (null hypothesis)
H1 or Ha
The two hypotheses must be mutually
exclusive of each other.
26
Ethics
 Talk about Milgram study and Zimbardo’s
study.
 ApA Ethics





The use of deception
Debriefing
Consent forms
Confidentiality
Regulation of research – IRB form – hand out.
27
Review of the Literature
 References
 Primary versus secondary sources
 Search terms
 Descriptors
 Professional journals – peer edited
28
Sampling
 Before we implement a research design,
there are a few more issues to discuss.
 One huge issue is that of sampling.
 Population
 Target population
 Sample
 Random sampling – Table of random
numbers
29
Different types of random sampling
 Systematic selection: here we have sampling
units in sequences separated on lists by the
interval of selection.
30
Stratified random sampling
 A separate sample is randomly selected
within each homogeneous stratum (or layer)
of the population.
31
Area Probability Sampling / or cluster
random sampling
 This is a type of stratification sampling
procedure.
 In this case the population is divided into
selected units that have the same probability
of being chosen as the unselected units in a
population cluster.
32
Two-stage random sampling
 Combine cluster random sampling with
individual random sampling
 For example… select 25 classes / 100
randomly selected.
 Then randomly select 4 students out of each
class.
33
Nonrandom sampling
 Convenience sampling
 Purposive sampling
34
Generalizability from a sample
 Population generalizability: the degree to
which a sample represents the population of
interest.
 Representative sample
 When random sampling is not feasible
35
Instrumentation
 Data
 Instrumentation: the design of the research
and procedures and conditions under which a
design is administered.
 Instruments used within the research design


Must have validity – measures what they are
suppose to measure
Reliability – give the instrument multiple times
– consistent results.
36
Buros Mental Measurements Yearbook
/ Subject completed Instruments
 Investigation of a previously researched
instrument
 Subject completed instruments:
Self-checklists (e.g. go into a counseling office
and check of current symptoms)
 Attitude scales
 Often given in likert scale design – e.g. 1-5,
1-7.

37
Types of tests
 Personality Tests


MMPI2/MMPIA – true/false format
Projective tests, e.g. TAT, Incomplete
sentences blank, Rorschach
 Achievement Tests

Woodcock Johnson (based on grade level)
 Aptitude Tests

Measures your ability to perform a task
38
Types of Scores
 Raw scores
 Derived scores – derived from a raw score
and used in a more standardized format

E.g. IQ tests – take the raw score and convert
numbers to a standardized score based on the
average of 100 IQ. Draw normal curve.
 Age and grade-level equivalents – child
compared to the age/grade level of the typical
level of performance, e.g. Woodcock Johnson
39
Norm-referenced versus criterionreferenced instruments
 Norm group – the group used to determine a
derived score.



For example … have a depression inventory.
Give the inventory to a norm group of clinically
depressed individuals.
Compare the items that clinically depressed
patients respond to with that of the subject.
Called a normed-referenced instrument.
40
Criterion referenced instrument
 There is a specific criterion for each person to
achieve.


The criterion for mastery or passing is fairly
high.
Example … Sylvan learning center …
41
Measurement Scales
 Nominal scale: Assign a number to different
categories.

E.g. 1=women; 2=men
 Ordinal Scale: data ordered in some way, e.g. rank
order students from highest grade to lowest.
 Interval scale: same as ordinal, but distances
between points are equal, e.g. temperature distances
between 0-10; 10-20 is the same.
 Ratio scale: an interval scale with a true zero point,
e.g. scale measurements like height. Zero=the
absence of height.
42