Psych220
a chill guy
chapter2.pptConducting Psychological Research
Slides Prepared by Alison L. O’Malley
Passer Chapter 2
1
What is “good science”?
Jot down 3 characteristics…
2
Origins of Research Questions
Personal experience and daily events
Prior research and theory
Real-world problems
Serendipity
Generate an example
associated with each
source.
3
Conducting a Literature Search
Where
to
begin?
4
Conducting a Literature Search
Online databases: PsycInfo, Google Scholar…
Boolean operators: AND, NOT, OR to narrow results
***Peer-reviewed articles***
Full text access?
If not, try authors’ personal websites
… or interlibrary loan (allow plenty of time!)
5
Conducting a Literature Search
Research Question: Are pet owners happier than non pet owners?
What’s the optimal What’s the optimal
Way way to enter this question into a search
database?
In search database?
| Manuscript component | Brief description |
| Abstract | Short summary of study |
| Introduction | Background and rationale for hypotheses |
| Method | Participants, procedure, materials/measures |
| Results | Data analysis – statistical tests reveal support or lack thereof for hypotheses |
| Discussion | Non-statistical review of findings, implications, limitations, avenues for future research |
| References | List of all in-text citations formatted in APA style |
Making Sense of What You Find
7
| Manuscript component | Brief description |
| Abstract | Short summary of study |
| Introduction | Background and rationale for hypotheses |
| Method | Participants, procedure, materials/measures |
| Results | Data analysis – statistical tests reveal support or lack thereof for hypotheses |
| Discussion | Non-statistical review of findings, implications, limitations, avenues for future research |
| References | List of all in-text citations formatted in APA style |
Making Sense of What You Find
8
| Manuscript component | Brief description |
| Abstract | Short summary of study |
| Introduction | Background and rationale for hypotheses |
| Method | Participants, procedure, materials/measures |
| Results | Data analysis – statistical tests reveal support or lack thereof for hypotheses |
| Discussion | Non-statistical review of findings, implications, limitations, avenues for future research |
| References | List of all in-text citations formatted in APA style |
Making Sense of What You Find
9
| Manuscript component | Brief description |
| Abstract | Short summary of study |
| Introduction | Background and rationale for hypotheses |
| Method | Participants, procedure, materials/measures |
| Results | Data analysis – statistical tests reveal support or lack thereof for hypotheses |
| Discussion | Non-statistical review of findings, implications, limitations, avenues for future research |
| References | List of all in-text citations formatted in APA style |
Making Sense of What You Find
10
| Manuscript component | Brief description |
| Abstract | Short summary of study |
| Introduction | Background and rationale for hypotheses |
| Method | Participants, procedure, materials/measures |
| Results | Data analysis – statistical tests reveal support or lack thereof for hypotheses |
| Discussion | Non-statistical review of findings, implications, limitations, avenues for future research |
| References | List of all in-text citations formatted in APA style |
Making Sense of What You Find
11
| Manuscript component | Brief description |
| Abstract | Short summary of study |
| Introduction | Background and rationale for hypotheses |
| Method | Participants, procedure, materials/measures |
| Results | Data analysis – statistical tests reveal support or lack thereof for hypotheses |
| Discussion | Non-statistical review of findings, implications, limitations, avenues for future research |
| References | List of all in-text citations formatted in APA style |
Making Sense of What You Find
12
| Manuscript component | Brief description |
| Abstract | Short summary of study |
| Introduction | Background and rationale for hypotheses |
| Method | Participants, procedure, materials/measures |
| Results | Data analysis – statistical tests reveal support or lack thereof for hypotheses |
| Discussion | Non-statistical review of findings, implications, limitations, avenues for future research |
| References | List of all in-text citations formatted in APA style |
Making Sense of What You Find
Note. Review papers (e.g., Annual Review of Psychology) will deviate from this format
13
Forming a Hypothesis
Inductive: Specific facts general conclusion
Data driven; “bottom up”
E.g., medical diagnosis based on symptoms
Deductive: General principle specific conclusion
Theory driven; “top down”
E.g., All people have ___. Pat is a person. Therefore, Pat has ___.
Is one logical approach “better” than the other?
REMEMBER: Above all else, hypotheses must be TESTABLE!!
14
Research Approaches: Key Distinctions
Describe the characteristics of a recent happy episode in your life.
How happy are you?
Qualitative vs. Quantitative
1 2 3 4 5
Research Approaches: Key Distinctions
Research Scenario 1: Employees randomly assigned to receive cookies or not receive cookies while completing a job satisfaction questionnaire (Brief, Butcher, & Roberson, 1995)
Research Scenario 2: Employees complete a questionnaire containing questions about mood and job satisfaction
Experimental vs. Descriptive
Research Approaches: Key Distinctions
Research Scenario 1: Employees randomly assigned to receive or not receive cookies while completing a job satisfaction questionnaire (Brief, Butcher, & Roberson, 1995)
Research Scenario 2: Employees complete a questionnaire containing questions about mood and job satisfaction
Experimental vs. Descriptive
What can we conclude on the basis of each research scenario? Why?
Research Design: Mind Your Variables
Independent variable: Systematically manipulated by the researcher in experimental research
Dependent variable: Outcome of interest; what we design research to assess/measure
Research Design: Mind Your Variables
Identify the IV(s) and DV(s) in this scenario:
Employees randomly assigned to receive cookies or not receive cookies while completing a job satisfaction questionnaire
Mastering IVs and DVs
Generate and describe a good strategy for distinguishing independent variables from dependent variables in research scenarios.
Research Approaches: Key Distinctions
Did employees complete the job satisfaction questionnaire under the same conditions (i.e., in identical environments), or did they take the questionnaire online at a time and place of their choosing?
Laboratory vs. Field
Lab settings = CONTROL
Research Approaches: Key Distinctions
Field experiments still entail manipulation of an IV, but occur in a natural setting as opposed to a lab setting.
Researchers often mention the tradeoff between internal and external validity. What exactly does this mean, and why does such a tradeoff occur?
Laboratory vs. Field
Research Approaches: Key Distinctions
Cross-sectional vs. Longitudinal
20 year olds
40 year olds
60 year olds
If all three age groups are measured and
compared in summer 2013, the design
is cross-sectional.
Research Approaches: Key Distinctions
Cross-sectional vs. Longitudinal
20 year olds
40 year olds
60 year olds
If all three age groups are measured and compared in summer 2013, the design
is cross-sectional.
Beware of cohort effects–different age groups
have different histories. Are observed
differences due to age differences or the groups’
different historical experiences?
Research Approaches: Key Distinctions
Cross-sectional vs. Longitudinal
20 years old Summer 2013
40 years old Summer 2033
60 years old Summer 2053
If a group of participants is measured repeatedly over time, the design is longitudinal.
Research Approaches: Key Distinctions
Cross-sectional vs. Longitudinal
20 years old Summer 2013
40 years old Summer 2033
60 years old Summer 2053
If a group of participants is measured repeatedly over time, the design is longitudinal.
Sequential research designs examine several age cohorts longitudinally.
Research Approaches: Key Distinctions
Cross-sectional vs. Longitudinal
20 years old Summer 2013
40 years old Summer 2033
60 years old Summer 2053
What are the advantages and disadvantages of longitudinal and sequential
research designs?
Research Design: Mind Your Variables
Internal validity is compromised by the presence of confounds, a particularly pesky type of extraneous variable.
Research Design: Mind Your Variables
Example: Do participants prefer stimuli associated with the first letter of the English alphabet?
If random assignment is used such that half the participants see the object on the left and half see the object on the right, what’s the problem?
A
B
The Role of Sampling
What is a population?
The entire group of scores that a researcher desires to learn about (e.g., all U.S. college students)
What is a sample?
A subset of scores from the population (e.g., 1,000 college students from a variety of colleges)
Population vs. Sample
Analyzing Data and Drawing Conclusions
Quantitative and qualitative analysis
Descriptive Statistics
Measures of central tendency address the typicality of a score:
Mode: most frequent score
Median: middle score (of an ordered set)
Mean: mathematical center of distribution
Organize and summarize a set of data
Descriptive Statistics
Build a dataset comprised of how many siblings each of your classmates has.
Establish the mode, median, and mean for this dataset.
Central Tendency
Descriptive Statistics: Central Tendency
Is it more appropriate to report the mean or the median for men and women in this dataset? Why?
MD = median
SP = sexual partners
Apologies for the fuzzy image
34
Descriptive Statistics: Measures of Dispersion
Measures of dispersion address the spread (i.e., the variability) of a set of scores.
Organize and summarize a set of data
Sketch the distribution associated with each of the three parties.
Descriptive Statistics: Measures of Dispersion
Measures of dispersion address the spread (i.e., the variability) of a set of scores.
Organize and summarize a set of data
Range: distance between highest and lowest score
Variance: spread of scores relative to mean
Standard deviation: square root of variance
Inferential Statistics
An oft heard question is whether research findings are “statistically significant.” Are our findings merely due to random error—to chance?
Inferential statistics reveal the probability that our findings are due to chance.
We use sample data to infer the nature of the population
Inferential Statistics
Psychological scientists traditionally maintain that findings are statistically significant if the probability is less than 5% that the results are due to random error.
We use sample data to infer the nature of the population
p < .05 =
Inferential Statistics: Drawing Conclusions
Statistically significant findings mean that we’ve proven how the world works, right?
We use sample data to infer the nature of the population
Inferential Statistics: Drawing Conclusions
Statistically significant findings mean that we’ve proven how the world works, right?
WRONG.
We use sample data to infer the nature of the population
Inferential Statistics: Drawing Conclusions
Our results may not be practically important…
…or perhaps there were confounding variables at play.
Good research design is critical!
And even with solid research design, maybe our conclusion is downright wrong.
We use sample data to infer the nature of the population
Drawing Conclusions
Two errors: False alarms and missed opportunities
An innocent person is found guilty False alarm (Type I error)
In research terms, we mistakenly conclude that two variables are associated
when they really have nothing to do with each other.
Drawing Conclusions
Two errors: False alarms and missed opportunities
A guilty person is found innocent Missed opportunity (Type II error)
In research terms, we mistakenly conclude that two variables are not associated
when they really are related.
Drawing Conclusions
Two errors: False alarms and missed opportunities
Apply the false alarm and missed opportunity
scenarios to the “cookie” experiment
(Brief et al., 1995).
How to Tell Your Research Story
So we all speak the same language!
Run, don’t walk, to access the 6th edition of the APA publication manual!
Helpful for students to see this is the most recent version
45
Building Knowledge and Theories
Contemplate the distinction between a theory and a hypothesis…
Now, why does theory building matter?
What Makes a Good Theory?
Testability and specificity
Does theory lend itself to testable hypotheses and specific predictions?
Internal consistency and clarity
Does theory avoid contradictory predictions? Can it be falsified? Is it clear to experts how components of the theory relate to each other?
47
What Makes a Good Theory?
Empirical support
Can theory be reconciled with current knowledge base? If not, can it debunk current “fact”? Does high quality research support new hypotheses derived from theory?
Parsimony
Law of parsimony: Explanations should use the minimum number of principles necessary to account for the maximum number of facts.
48
What Makes a Good Theory?
Last, but not least: Does the theory have an impact on the field?
Proof and Disproof
Science values lively debate. There is no tolerance for the notion of “absolute proof.” It’s always possible that our results are due to chance. Similarly, a single set of results cannot “disprove” a hypothesis derived from a theory.
Science is forward-moving, and theories are strengthened or weakened as supportive or unsupportive findings continually emerge.
Research is more “probabilistic” than “absolute” (Baumeister, 2008)
63946.418
