Experimental Exam 2

Chapter 1

1.1 How Psychologists Acquire Knowledge

Studies of behaviors

- animals

- people

Research methods

- Research question dependent

What do you want to know?

- Wide variety of methods

Which one to use depends on the research question you want to answer

Quiz Questions

1. How do psychologists know things?

- Systematic study using scientific method

2. What drives the development of a research design?

- The research question you’re trying to answer

3. How do you know what research method to use?

- Pick the method that will best answer your question

1.2 The Research Process

(1 of 11) • Finding a Problem

- Beginning of a project (e.g., Not all students enjoy classroom success)

What variables lead to higher levels of achievement?

 (2 of 11) • Reviewing the Literature

- Explore research studies – what have other researchers already found? (e.g., Deci & Ryan, Elliot, Dweck)

 (3 of 11) • Theoretical Considerations – what theoretical framework will you adopt?

- Definition of theory (e.g., SDT, achievement goals, and implicit theories – pick one!)

- Two common properties of all theories

» Attempt to organize existing data

» Lead to new research

 (4 of 11) • Hypothesis – Attempt to state specific relationships within a larger area or theory

- Multiple hypotheses for a research question (e.g., students with higher SDT, mastery goals, and incremental implicit theories will learn more and earn higher grades

 (5 of 11) • Research Plan (Research design)

How will you carry out your study to answer your research question? (Research question? Participants? Quall or Quant? Experimental or Correlational? Instruments? (This is what you use to write your IRB application)

(6 of 11) • Conducting the Research Project

- Location

- Participants

- Procedures

e.g., DV=Classroom success, IV=achievement goals

 (7 of 11) • Analysis of Research Findings

Data Analysis

Qualitative – findings emerge as you study data, looking for meaningful patterns and themes (not


Quantitative – examine numerical data results (could be experimental)

- Statistically significant?

- Substantively significant?

 (8 of 11) Discussion – what did you find and what does it mean?

• Decisions in Terms of Past Research and Theory

- Results interpretation

- Generalizability

 (9 of 11) • Preparing the Research Report

- APA formatting

e.g., Proposal assignment

1. Introduction (background to problem)

2. Literature Review (conclude with purpose of study and research question)

3. Method (Research design, N, methodology)

4. Results (what you found)

5. Discussion (conclusions & recommendations)

 (10 of 11) • Sharing Your Results: Presentation and Publication

- Psychological conventions

- Journal publication

 (11 of 11) • Finding a New Problem

- recommendations for future study

What new questions emerged as you answered the old one?

Quiz Questions

4. What’s in the Introduction?

- Background to problem

5. What’s in the Lit Review?

- Summary of existing relevant research

6. How do you conclude the Lit Review?

- With purpose of study & research question

7. What’s in the Method section?

- Research design, N, methodology

8. What’s in the Results section?

- what you found

9. What’s in the Discussion?

- conclusions & recommendations

1.3 Why Is the Research Methods Course Important?

- Assisting you in other psychology courses

- Conducting an original research project

- Conducting a research project after graduation

- Getting into graduate school

- Becoming a knowledgeable consumer of research

Quiz Questions

10. Why should I learn this stuff? (besides it’s required)

• To be a better psychology student

• To become a researcher yourself

• To be better prepared for life after graduation

• To be able to evaluate other people’s research and determine for myself if I think it’s valid

Chapter 2

2.1 The Research Idea

(1 of 2) • Characteristics of Good Research Ideas

- Is it testable?

• Not all questions can be answered now

• How will you test it?

- What is the likelihood of success?

• Design a project as close to reality as possible

• Start with past research

 (2 of 2) • Sources of Research Ideas

- Nonsystematic sources

• Unpredictable

• Inspiration

• Serendipity

• Everyday occurrences

- Systematic sources

• Past research

• Theory

• Classroom lectures

2.2 Developing a Research Question

- Convert idea into a question

• Guides scientist (you) through research ideas

• Leads to the literature review

Quiz Questions

1. How do you know if you’ve come up with a good research idea?

If it’s testable and has a good chance of working

2. Where can you get good research ideas?

Past studies, current theories, and coursework or inspiration from everyday life

3. What’s the value of turning your idea into a research question?

It keeps you focused to design a good study

2.3 Surveying the Psychological Literature

Critically evaluate all Internet sources (hint: use library)

Selection of index terms (what are you specifically interested in?)

• Google Scholar

• Ebsochost (E.g., PsycINFO and PsycArticles)

• Sage

Obtain relevant publications

• Library

• Interlibrary loan

• Request to author

Consult with information specialist professionals(e.g., librarians) for cutting edge research results

For simpler data management (so all your hard work doesn’t go to waste) —

Integrating the results of the literature search to include

• Reference information (for citation)

• Introduction (what was the problem?)

• Method (how did they study it?)

• Results (what did they find?)

• Discussion and evaluation (what did it mean?)

Condense into a single page (helps you internalize it better and access it later)

Quiz Questions

4. What’s the best way to make sure information sources are credible?

- Consult the librarians

5. What information do you need to synthesize from each source for future use, and why?


- Citation information for reference page

- Problem in the intro

- How they studied it in the method

- What they found in the results

- What they think it meant in the discussion


- So you can use it later

2.4 The Need for Ethical Principles

- 1974 National Health Research Act

• Federally funded projects need approval

• Human Subjects Review Panel

• Institutional Review Boards

- World War II

• Nazi research (viruses, toxins, drugs)

• Nuremberg War Tribunal

• Nuremberg Code (participants must be informed, give consent, be protected, be free from undue risks and be free to stop at any time; researchers must be qualified scientists)

- Willowbrook (institution for patients with intellectual disabilities)

• Hepatitis research

- Tuskegee syphilis study

• 399 African American men

• “Bad blood”

• Lack of treatment

- Milgram’s obedience-to-authority study

• Cause of intentional distress

• Debriefing

Quiz Questions

6. What are the stipulations of the Nuremberg Code and why was it developed?


a. must be informed

b give consent

c. be protected

d. be free from undue risks

e. be free to stop at any time

Researchers must be qualified scientists

•Nuremberg Code was developed after WWII in response to unethical Nazi research on prisoners

7. Name the primary ethical violations of the Willowbrook, Tuskegee, and Milgram’s studies

•Willowbrook – intellectually disabled clients were injected with hepatitis (parents felt coerced)

•Tuskegee – deceived African American men with syphilis to think they were being treated for “bad blood” and denied treatment for their illness

•Milgram – deceived participants to think they had followed orders to severely shock other participants, which induced extreme emotional distress

2.5 APA Principles in the Conduct of Research With Humans

• Ideally, we want to provide a complete explanation of the research in advance

• Is Deception in Research Necessary?

• Sometimes a complete explanation can influence results

- Justified in some cases

• When participants are likely to “figure out” the “right” response

• When participants think the researcher expects certain responses

• Informed Consent

- Consent form

• Beginning of study

• General description

• No penalty

• Right to withdraw

• Informed Consent

- Children

Multiple levels of approval

- Deception

• General statement

• Placing people at risk

• Participants at Risk and Participants at Minimal Risk

- Participants at risk

• Emotional (e.g., trauma survivors)

• Physical (e.g., children, the disabled)

• Mandatory informed consent in every case

• Participants at Risk and Participants at

- Participants at minimal risk

• No risk of harm (e.g., public behavior)

• Not mandatory to obtain informed consent

- Participants at risk and deception

• Inform participants of potential harm

• Debrief afterward

• Vulnerable Populations

- Special considerations

• Poor health

• Age

• Children

• Mentally disabled

• Physically disabled

- Call for specific guidelines

• The Debriefing Session

- Demonstrate Integrity as a scientist

Researcher’s dedication to scientific method

- Deception explained

Reassure participants

- No rushing!

- Sensitive to discomfort

Return to emotional state prior to study

- Confidentiality and anonymity

Believable assurance

- Immediate debriefing

No delay

Quiz Questions

8. When might deception be warranted, and what steps must be taken if deception is used?

• When the participants can easily “figure out” the correct response or when they feel the researcher expects a certain response

• Informed consent must be obtained

• Quality debriefing must be provided afterward

2.6 The Ethical Use of Animals in Psychological Research

Debates of animal research

Animal Welfare Act of 1966

• Humane treatment of animals

APA requirements

a. Comply with regulations and professional standards

b. Trained to ensure comfort, health, and humane treatment

c. All support staff members properly trained for their roles

d. Never do harm unless scientifically justifiable and no alternatives exist

e. Surgery must include anesthesia and after-treatment health care

f. If life is to be terminated, this must be carried out humanely to minimize pain and in accordance with accepted procedures

2.7 The Institutional Review Board

- Composition

• Cross-section of individuals

• Peers

- Purpose

• Not scientific merits

• Methodology

• Informed consent

• Risk

- Forms

• Institutional

Quiz Questions

9. What’s an IRB, and what is its purpose?

•The Institutional Review Board is a cross section of professional peers

•Evaluates the planned research design including method, informed consent, and risk, to safeguard the ethical treatment of human and animal subjects

2.10 The Researcher’s Ethical Obligations Once the Research Is Completed


When using a direct quote (i.e., Exact words)

• Quotations

• Author information

• Year of publication

• Page numbers

Minor changes

• Avoid word substitution (still plagiarism)


• Citations still necessary

Secondary sources (try to avoid!)

• At minimum, acknowledge by citing appropriately: Erikson (1964, as cited in Bergen, 2011) posited a series of psychosocial stages…

• Citing Your References Correctly

- Secondary sources

• When primary truly cannot be located. E.g., personal correspondence from Allport to James:

Allport (as cited in Jones, 1999)…


Reference facts and opinions

• Cite unless general knowledge for your reader

Repeat papers

• Cannot turn in a paper from another course

Proofreaders (peer reviewers)

• Criticism is fine; do not use others’ work

Keep materials

• Rough drafts and notes

Fabrication of Data

- Deliberate alteration of data

• Ethical violation

• Burt’s twin studies

• Lying With Statistics

Be careful to include sufficient detail in Method and Discussion sections to describe exactly what you did and why, so readers can assess validity for themselves

- Individual choice

• Creativity 

• Questionable practices

Quiz Questions

10. What are the two main ways researchers fail to meet their ethical obligations?

•Fail to cite others’ work appropriately

•Fabricate/distort data

Chapter 5

5.1 Components of the Scientific Method

• Objectivity

- Empirical

Bias: a tendency to believe that some people, ideas, etc., are better than others that usually results in treating some people unfairly; a strong interest in something or ability to do something

• Confirmation of Findings

- Replication (confirm previous findings)

- Replication with extension (confirm with new findings)

• Self-Correction

- Errors and faulty reasoning discovered

- New conclusions are drawn and/or old conclusions are refined

• Control

-variables are kept constant to improve accuracy of the experiment

Quiz Questions

1. What are four features of a scientific study in


- Objectivity based on empirical evidence, not biases

- Confirmation (or disconfirmation) of past findings

- Self-correction as theories are tested and subsequently proven, disproven, and/or refined

- Controls imposed on variables to improve accuracy of findings (to ensure we’re studying what we think we are)

5.2 The Psychological Experiment

Independent Variable: I can change it.

What changes in an experiment?

E.g., which type of cleaning product removes dust best? We’ll test 3 different kinds (3 levels)

IV = kinds of cleaning product

Dependent Variable: Depends upon the independent variable

-what I’m interested in; what am I trying to study?

E.g., the clean, dust-free quality of the object that was cleaned by the different products tested

DV = amount of dust

• Extraneous Variables

- other influences upon DV

Quiz Questions

2. What are the three main types of variables in a psychological experiment?

- Independent variable that the investigator manipulates to measure its effect on the dependent variable

- Dependent variable that the investigator measures to determine the effect (if any) of the independent variable

- Extraneous variables other than the IV that may exert an influence over the DV

5.3 Establishing Cause-and-Effect Relations

Discovering cause and effect – The purpose of the experiment

• Why does this happen?

• Requires control of extraneous variables

• Requires random sampling

Quiz Questions

3. What’s the purpose of a psychological experiment?

- To establish a cause and effect relationship between the independent and dependent variables

5.4 Formulating the Research Hypothesis


• Helps organize data

• Suggests IV-DV relationships

- States your a priori prediction

May contribute to theory, if supported

Quiz Questions

4. Why do we need to formulate a hypothesis?

- Helps to organize our data

- Suggests the IV-DV relationships

- Articulates our prediction of what the research will show us

5.5 Characteristics of the Research Hypothesis

Types of Statements

- Synthetic

• True or false statements

• Research hypotheses

• “I understand my notes and I will make an A.”

- Analytic

• Always true statements

• “I’m making an A or I’m not making an A.”

- Contradictory

• Always false statements

• “I’m making an A and I’m not making an A.”

Types of Statements

- General implication form

• If…then

• If = IV

• Then = DV

• Principle of falsifiability

Quiz Questions

5. What kind of statement should a hypothesis be and why?

-Synthetic statements because they can be proven to be either true or false.

6. What’s General Implication Form and why is it helpful to structure a hypothesis?

-If…then statements structure the hypothesis into a testable synthetic statement: If the IV… then the DV…

5.5 Characteristics of the Research Hypothesis

Types of Reasoning

- Inductive logic

• Reasoning from specific cases to general principles

• Construction of theories

• Ex.: Kitty Genovese and bystander effect

Types of Reasoning

bystander effect

- Deductive logic

• Reasoning from general principles to specific cases

• Formulating research hypotheses

• Research uses both types of logic

Quiz Questions

7. What’s the difference between inductive and deductive logic?

-Inductive logic is reasoning from specific cases to a general conclusion (“We each understood our notes and we all made a good grade.”)

-Deductive logic is reasoning from general beliefs to specific cases (“If I understand my notes, then I will make a good grade.”)

8. Which kind of logic is used in research?

-Both kinds of logic are used in research to theorize, test, and refine our understanding about why things are the way the way they are (“Understanding your notes leads to higher grades.”)

5.5 Characteristics of the Research Hypothesis

• Directional Versus Nondirectional Research Hypotheses

- Directional research hypotheses

• Specify outcome of experiment

• When relatively certain of outcome

• Directional Versus Nondirectional Research Hypotheses

- Nondirectional research hypotheses

• Predicts change, but not direction

• Most researchers choose this type

Quiz Questions

9. What’s a directional hypothesis?

- A synthetic statement that statement that states the specific nature, or direction, of the IV’s effect on the DV (A happy dog will eat more.)

10. What’s a nondirectional hypothesis?

- A synthetic statement that predicts the IV will have an effect on the DV but not the specific nature, or direction, of this effect (Mood affects appetite in dogs.)

5.5 Characteristics of the Research Hypothesis

An Alternate View of Hypothesis Testing

- Hypothesis testing is complicated

• Importance of research problem

• Can be harmful

• Incorrect auxiliary assumptions

• Use inductive logic when research area is new

Chapter 6

6.1 The Nature of Variables

Definition: an event or behavior that can assume at least two values

• Temperature

• Height

• Anxiety

• Self-esteem

- Applies to IV and DV

- Strive to avoid extraneous variables with two or

more values

• Presenter

• Climate

Quiz Questions

1. What property must be present in a psychological variable?

Both independent and dependent variables must be able to assume at least two values.

6.2 Operationally Defining Variables

Replication of past research – good source of research ideas

- Operational definitions: variables should be defined in terms of the operations required to produce them

• Depression>BDI

• Loneliness>UCLA Loneliness Scale

- Clearly communicates what is being studied and how it is being measured

Quiz Questions

2. What’s an operational definition, and why is it important?

Defines each variable studied in terms of the operation required to produce it, communicating precisely—

• what is being studied

• how it is being measured

6.3 Independent Variables

Types of IVs

- Definition: manipulated (I control them)

(If…then, the IV is described by If)

• Physiological – biological state (e.g., alcohol levels)

• Experience – learning effects (e.g., memory cues))

• Stimulus – environmental (e.g., high heels v. bare feet)

• Participant – individual characteristics (commonly used, but not a true IV, because this cannot be manipulated/randomly assigned by the researcher) classification – age, sex, personality, etc.

Quiz Questions

3. What are four different types of IVs, and which one is not a true IV?

• Physiological

• Experience

• Stimulus

• Participant (not a true IV because the researcher cannot manipulate it)

6.4 Extraneous Variables (Confounders)

Definition and effects

• Factors that can have unintended effects (influence the difference between groups)

-Devastating to research because we can’t know if differences between groups are due to the IV or the confound

e.g., two classes with two different instructional styles being tested for effectiveness

Confound = different teachers

Is the difference between groups attributable to the instructional styles of the teachers?

Quiz Questions

4. What’s a confounding variable and how does it affect the DV?

A variable besides the IV that affects the DV and creates between-group differences, confounding findings.

6.5 Dependent Variables

• Selecting the DV (or the variable that depends on the IV for change; often a behavior or response)

“If…then” (the DV is described by then)

- Choosing the dependent variable

• Remember the variable must be operationally defined

• Replication is a good place to start

Recording or Measuring the DV

- Correctness/accuracy (answers correct)

- Rate or frequency (aggressive acts at recess)

- Degree or amount (LSS scores)

- Latency or duration (how quickly or how long)

e.g., latency could be reaction time; duration could be tantrum length

Recording More Than One DV

- When to record more than one DV

• First question: do you have the measurement capabilities?

• Second question: will it add to the study?

e.g., mirror tracing test

How long it takes to complete a star tracing task

Duration is important, but correctness also matters

• Characteristics of a Good DV

- Valid (measures what it’s supposed to)

- Reliable (measure is consistent – are the scores highly similar in a test-retest procedure?)

Quiz Questions

5. What’s are the characteristics of a good DV, and what are four ways to measure it?

A good DV should be valid and reliable and could be measured in terms of the following:

- Correctness/accuracy

- Rate or frequency

- Degree or amount

- Latency or duration

6.6 Nuisance Variables

What are these??

Similar to a confound, but they’re not usually limited to one group, so they tend not to differ between levels of IVs (i.e., don’t cause between group differences)

Rather, they increase variability within groups to affect scores on the DV

e.g., DV = reading comprehension

What are some participant characteristics within all groups that could affect this score?

Intelligence? Verbal ability? Education/experience?

Quiz Questions

6. What’s a nuisance variable and how does it differ from a confounding variable?

A nuisance variable is usually due to within group differences that affect the DV and does not show up as differences between groups like a confound.

6.7 Controlling Extraneous Variables

• Basic Control Techniques

- Randomization

- Elimination

- Constancy

- Sequence or order effects

- Carryover effects

- Balancing

- Counterbalancing

Randomization (random assignment)

• Most widely used

• Each participant has equal chance of being assigned to a group

• Unique characteristics equally distributed

• Never fully aware of the variables and if they are equally distributed or not

Quiz Questions

7. What’s the most commonly used control for extraneous variables, and why is it effective?

The most common control is random assignment, which is effective because it spreads unknown extraneous variables evenly across groups, so they are less likely to skew findings unevenly

6.7 Controlling Extraneous Variables


• Remove unwanted variable(s) (e.g., if the variable is faces and you don’t want participants distracted by hair, clothing, or body type, you might crop the picture so only the faces showed)

• Difficult to completely remove (e.g., noise, temperature, etc.)


• Create uniform condition

• Also controls for nuisance variables

• Not all variables can be reduced to a single value; an option can be to hold the extraneous variables constant

(e.g., one lighting condition, location, temperature, teacher, sex studied, etc.)

Quiz Questions

8. What are two common techniques to control extraneous variables when the researcher knows what they are?

-Eliminating the extraneous variable

-Holding the extraneous variable constant across groups/conditions

6.7 Controlling Extraneous Variables

Sequence or order effects

• Participant’s exposed to sequential presentation of the treatments

E.g., reaction time: turn off engine when presented with warning light stimulus (red, green, yellow)

Order effect-1st=10 sec, 2nd=4 sec, 3rd=3 sec

• Controlled in counterbalanced situations (but not with incomplete counterbalancing)

Carryover effects

• Effects of one treatment continue to influence the participant’s response to the next treatment

E.g., taking a depression inventory prior to a life satisfaction inventory could lower scores on the second measure; the reverse could be true if the order of administration changed

Quiz Questions

9. What are order effects and carryover effects?

An order effect occurs when the sequence in which a treatment is administered affects the DV.

A carryover effect is when the treatment itself affects the DV.

6.7 Controlling Extraneous Variables


An extension of constancy—

• All unwanted variables and levels of unwanted variables treated in the same manner

• Two groups treated identically, except for manipulation of IV

(e.g., two instructional methods and a control with two teachers – balancing would require each teacher to teach a class with each method)

Counterbalancing (e.g., taste test)

• Within subject (each person tastes each soda in every possible sequence – ABBA)

• Within group (randomly assign ½ participants to one sequence and the other ½ participants to the other)

n! = take the number of treatments (n), factor into component parts, then multiply

E.g., 2 sodas

2! Is 2×1=2 groups

E.g., 4 sodas

4! Is 4x3x2x1=24 groups

6.7 Controlling Extraneous Variables

Incomplete counterbalancing

Possible solution when insufficient groups are available

• Use some, but not all, of the possible sequences

• E.g., randomly select the sequence for the first participant than systematically rotate the sequences for remaining subjects




Quiz Questions

10. What is balancing, counterbalancing, and incomplete counterbalancing?

Balancing creates equivalent groups by assigning all extraneous variables evenly to all groups.

Counterbalancing eliminates within-subject or within-group differences by ensuring each treatment is given to each subject in each possible sequence an equal number of times

Incomplete counterbalancing is similar to counterbalancing but not all possible sequences are used.

Chapter 7

7.1 Types of Participants

- Precedent

• Prior research

• Willard Small (1901) and rat study

• College students

• Higher likelihood of success

• Limited generalizability

- Availability

• Rats are inexpensive

• College students are accessible

• Related to precedent

- Type of the problem

• Nature (animals)

• Preschoolers

• Specific to problem

Number of Participants

- Practical considerations

• Finances

• Time

• Availability

- Expected variability in each group

• Greater heterogeneity = larger number

• Stratified random sampling

- Statistical test

• Larger sample = more powerful the test

- Literature review

• Prior research

Quiz Questions

1. What variables affect the type and number of participants needed?

• The type of participants chosen depends upon precedent, availability, and the problem under study (what’s the question?)

• The number of participants recruited depends upon practical concerns (e.g., finances, time, and availability), how much variability is expected within the group, the statistical methods you plan to use, and precedent

7.2 Apparatus

• IV Presentation

Nature of IV determines apparatus

For example—

• Audio equipment

• Skinner box

• Creativity

Quiz Questions

2. How do you determine the best way to present the Independent Variable?

• The apparatus needed to present the IV depends upon the nature of the IV. (I.e., what is the IV you need to manipulate, and what’s the best way to do so?)

7.2 Apparatus

• DV Recording

- Reactivity

• Researcher presence

- Technology

• Over-reliance on equipment

• Keep it simple

Quiz Questions

3. What factors should be taken into account when deciding how to present the Dependent Variable?

• When determining how to present the DV, you need to consider the affect the researcher’s presence might have and what technology might be needed.

7.3 The Experimenter as an Extraneous Variable

• Experimenter Characteristics

- Physiological

•Age, sex, race

- Psychological

•Hostility, extraversion, etc.

Experimenter Expectancies

- Cause of desired/expected response has influenced results in studies with people and animals

• Confounding variable

Ex.: Teacher expectations/IQ tests

• Rosenthal effects

Quiz Questions

4. How can the experimenter/researcher be an extraneous variable?

• The experimenter can have physical or psychological characteristics that affect participants.

• The experimenter can produce either the “good participant effect” or the “Rosenthal effect”

7.3 The Experimenter as an Extraneous Variable

• Controlling Experimenter Effects

- Standardized methods

• Instructions

• Scoring

• Mannerisms/demeanor

- Instrumentation/automation

• Tape recording

• Computer display

• Printed directions

- Single-blind experiment

• Experimenter unaware of who is in which group

Quiz Questions

5. What are some ways to control for experimenter effects?

• Standardization in instructions, procedures, scoring, and even presenter demeanor/affect

• Automatize instrumentation (tape recordings, computer display, printed directions)

• Single blind experiment (experimenter doesn’t know which treatment is administered to whom)

7.4 Participant Perceptions as Extraneous Variables

• Demand Characteristics and Good Participants

- Demand characteristics

• Participants may try to figure out what the experiment is about

• Try to act a certain way

• Good participant effect

• Extraneous or nuisance variable

Quiz Questions

6. What is meant by “demand characteristics” and “the good participant effect”?

• Demand characteristics refer to the characteristics of the experiment that hint to the participant what the researcher’s hypothesis is (i.e., what the researcher hopes to prove).

• The good participant effect is observed when the participant wants to comply with the researcher’s agenda to “prove” the hypothesis.

7.4 Participant Perceptions as Extraneous Variables

• Response Bias

- Yea-saying

• Yes to all questions

• Alternatively: Naysayers

- Response set

• Experimental context

Ex.: Interview

• How questions are worded

Quiz Questions

7. What is “response bias,” and what are some examples?

• Response bias is when participants are biased to respond in a given manner.

• Yea-saying is when the participant is inclined to answer every question affirmatively (and naysaying is the opposite case).

• Creating a response set is when the cues given along with the question prompt a certain kind of response as socially desirable for the setting.

7.4 Participant Perceptions as Extraneous Variables

• Controlling Participant Effects

- Demand characteristics

• Double-blind experiment

• Deception

- Yea-sayers

• Rewrite questions

• Order of presentation

- Safeguarding against the response set

• Carefully review all questions to avoid socially desired responses

• Pilot testing        

• Control for impression management

Quiz Questions

8. What are some ways to control for participant effects?

• Demand characteristics – use double blind experiment (neither experimenter nor participant knows which treatment is administered to whom) and consider whether it’s appropriate to mislead or use deception

• Yea-saying – rewrite some questions and/or vary the order of the questions presented

• Response set – Edit questions so that responses are most likely to reflect true feelings and not a socially desirable or expected response; consider pilot testing to explore this risk

7.5 The Interface Between Research and Culture

• Culture, Knowledge, and Truth

- Etic: Universal truth

- Emic: Culture-specific

• More emics than etics  

• E.g., Fundamental attribution error

• The Effect of Culture on Research

- Choice of the research problem

• Culture dependent

• Ex.: Rock concerts in United States vs. Australian bush

- Nature of the experimental hypothesis

• Cultural difference affects hypotheses

• Ex.: Personal space

- Selection of the IV and recording of the DV


• Ex.: Computers in developed nations

Quiz Questions

9. What are some problems that arise when culture isn’t taken into account in designing a research study?

• Not all findings represent absolute truths consistent across all people groups, leading to flawed methods and fallacious conclusions.

• Culture exerts a powerful influence and should be considered throughout the research process from the conception of the research question to the interpretation of findings.

7.5 The Interface Between Research and Culture

• Methodology and Analysis Issues

- Participants and sampling procedures

• Representative of the culture

• Ensure samples are equivalent

- Type of survey or questionnaire used

• Translations

• Back translation

• Value of concept being measured

• Specific items are equivalent

- Cultural response set

• Tendency of a culture to respond a certain way

• Ex.: Likert-type scales and familiarity with measure

• Culture as confounder

Quiz Questions

10. What are some specific methodological issues that should be considered in conducting culturally valid research?

• Participants – Is the sample representative of the population, and are comparison samples from different cultures truly equivalent?

• Instruments – Are written materials correctly translated?

• Response set – Is there a tendency for a cultural group to respond in a certain way, and is the method chosen familiar/well-understood by all participants?

Chapter 8

8.1 Internal Validity: Evaluating Your Experiment From the Inside

• Threats to Internal Validity

- History: Any significant event, except for IV, that occurs between DV measurements Ex.: Nazi propaganda; Distractions

- Maturation: Systematic time-related changes (Boredom, exhaustion); More likely in experiments with repeated measurements of DV

- Testing: Practice effect Ex., ACT/SAT

- Instrumentation: Malfunction; Interrater reliability

- Statistical regression: Extreme scores and regression to the mean; Premeasurements

- Selection: Select equivalent groups; Group membership

Quiz Questions

1.What are some threats to internal validity?

- History – big or small events

- Maturation – aging, improving, or showing lack of motivation (validity drift)

- Testing – practice effects and reactive measures

- Instrumentation – accurate, reliable, and in good


- Statistical regression – regression to the mean

- Selection – recruiting participants to be sure groups are truly equivalent (different from group assignment)

8.1 Internal Validity: Evaluating Your Experiment From the Inside

• Threats to Internal Validity (cont’d) - Mortality: animals-death, humans-dropouts

- Interactions with selection: Groups selected show differences on another variable (i.e., maturation, history, or instrumentation) that vary systematically by groups Ex.: Language development study

- Diffusion or imitation of treatment: Minimizes differences between groups Ex.: Studies of learning and memory – People pass information along to other groups

Quiz Questions

2. What are other threats to internal validity?

- Mortality – death or dropping out

- Interactions with selection – when groups we select vary systematically on another variable

- Diffusion or imitation of treatment – all groups “figure out” the treatment and respond accordingly

8.1 Internal Validity: Evaluating Your Experiment From the Inside

• Protecting Internal Validity

- Implement control procedures: Developed to help control problems

- Standard procedure: Experimental designs • Importance of internal validity

Internal Validity is the most important property of experiment

- Can’t establish cause and effect without determining changes in DV are due to IV

Quiz Questions

4. How can you protect internal validity?

-Use controls

-Follow standard procedures

5. How important is internal validity?

-Internal validity is the most important property of any experiment

6. Why is internal validity so important?

-Internal validity is the property that allows you to establish a cause-and-effect relationship. An experiment without internal validity is useless.

8.2 External Validity: Generalizing Your Experiment to the Outside

• Population generalization – findings apply to all populations

• Environmental generalization – findings apply in all environments

• Temporal generalization – findings apply at all times

Quiz Questions

 7. What are the three customary types of generalization?

• Population generalization – findings apply to all populations

• Environmental generalization – findings apply in all environments

• Temporal generalization – findings apply at all times

8.2 External Validity: Generalizing Your Experiment to the Outside

• Threats to External Validity (Based on Methods)

- Interaction of testing and treatment: Pretest–posttest control group highest threat to external validity; Pretest has effect on participants

- Interaction of selection and treatment: Effect holds true for only one group ; Risk increases with difficulty in finding participants

- Reactive arrangements: Artificial atmosphere of experiment; Behaviors affected by artificiality Ex., Hawthorne studies; Demand characteristics — gives clues on how to behave

- Multiple-treatment interference: Repeated measures designs Ex., Memory studies

Quiz Questions

 8. What are some threats to external validity based on Methods?

- Interaction of testing and treatment: Practice effects for control group

- Interaction of selection and treatment: Effect holds true for only one group

- Reactive arrangements: E.g., Hawthorne studies

- Multiple-treatment interference: Repeated measures designs e.g., Memory studies

Threats to External Validity (Based on Our Participants)

- The infamous white rat: Large number of studies using the Norway rat ; Limits generalizability -rats not close to humans

- The ubiquitous college student: Convenience sample; coercion; representative of all humans?

- The “opposite” or “weaker” or “inferior” or “second” sex: Women ignored Ex., Freud, Erikson; presumption that males are the “norm;” Caution for generalizability

Threats to External Validity (Based on Our Participants) (cont’d)

- Even the rats were white: Minorities ignored; presumption that Caucasians are the “norm”

- Even the rats, students, women, and minorities were American: Discipline became Americanized; presumption that Americans are the “norm;” 1960s saw increased awareness of culture; Cross-cultural psychology

Quiz Questions

 9. What are some threats to external validity based on Participants?

- Precedent (e.g., white rats are not human)

- Availability (e.g., college students are not representative of everyone)

- Assumptions about who is normal (e.g., sexism)

- Assumptions about who is normal (e.g., Euro- centrism)

- Assumptions about who is normal (e.g., nationalism – most psychological research has been set in the US)

8.2 External Validity: Generalizing Your Experiment to the Outside

• The Devil’s Advocate: Is External Validity Always Necessary?

- Douglas Mook: Generalizability not always intended or meaningful

Ex., Harlow’s work with baby rhesus monkeys

External validity only necessary when trying to predict real-life behavior in the real world

Four alternative goals

- Controlling for external validity: May be impossible

1. Control for interactions, reactive arrangements, multiple treatments

2. Careful experimental planning

Participant-related threats-”not logically solvable in any neat, conclusive way”

3. Replication

4. Replication with extension

 Quiz Questions

10. If you’ve established internal validity, and if you’ve controlled for interaction of testing and treatment, reactive arrangements, and multiple- treatment interference, what is one powerful way to address threats to generalizability based on participants?

- Test for invariance through multiple studies. It is shortsighted to test continually the same types of participants in every experiment. Replication with extension to retest for a particular experimental finding in a different context and/or with different participants should improve external validity and lead to more generalizability.

Chapter 10

10.1 Experimental Design: The Basic Building Blocks

The Two-Group Design

- The principle of parsimony (Occam’s Razor) Don’t needlessly complicate the question

- How many IVs? (one or more)

- How many groups? (“levels;” at least two)

- Assigning participants to groups (random or not?)

Comparing Two-Group Designs Goal: Maximize between group differences

Minimize error variance

Statistic=between group variability/error variability

- Choosing a two-group design

• Large groups — typical procedure is random assignment

• “Large” varies per researcher/situation

Comparing Two-Group Designs

Why do we favor random assignment?

Is there ever a more precise way to do this?

- Advantages of correlated-groups designs (within-subjects differences)

• (matched pairs-one for each group sharing a potentially influential characteristic: sex, politeness, intelligence, etc.)

• We exert more control to assure equal groups (rather than leaving it to random chance)

Statistic=between group variability/error variability

(individual differences are taken out of the error term>bigger t)

 Copyright © 2016 Pearson Education, Inc. All rights reserved.

10.1 Experimental Design: The Basic Building Blocks

Comparing Two-Group Designs

Why do we favor random assignment?

Is there ever a more precise way to do this?

- Advantages of independent-groups designs (between- subjects differences)

• Simpler and sometimes only option

• More degrees of freedom (smaller critical value to achieve significance)

e.g., 2.1 critical value for independent v. 2.2 for dependent

 Copyright © 2016 Pearson Education, Inc. All rights reserved.

10.1 Experimental Design: The Basic Building Blocks

• Comparing Two-Group Designs

- Advantages of independent-groups designs (between- subjects differences)

• More degrees of freedom (smaller critical value to achieve significance)

• Statistic=between group variability/error variability

• 20 subjects in independent v. matched pairs design

• (N-1)+(N-1) = 18 df v. (N-1) = 9 df

However, this tends to be a smaller difference as the sample size gets larger

 Copyright © 2016 Pearson Education, Inc. All rights reserved.

10.1 Experimental Design: The Basic Building Blocks

Variations on the Two-Group Design

- Comparing different amounts of an IV • Presence-absence of a manipulation (Control v. treatment)

• Different amounts of types of an IV

(Small amount v. larger amount) – Dealing with measured IVs

• Designs are not true experiments when you can ́t randomly assign IVs (e.g., sex, anxiety levels, etc.)

• Ex post facto research (quasi-experimental)

10.2 Statistical Analysis: What Do Your Data Show?

The Relation Between Experimental Design and Statistics

- Link between design and statistics

• Determine design first

• Ensure there are statistics appropriate for the design

Analyzing Two-Group Designs

- t tests

• Two-independent-groups design = t test for independent samples (independent t test)

• Two-correlated-groups design = t test for correlated samples (dependent t test)

• t test is appropriate for an experiment with one IV with two levels

Calculating Your Statistics

- Which statistical test is appropriate?

• One IV, two levels (or groups) – T test

• Random assignment and Independent t test

• Repeated measures/matched pairs and Dependent t test

*Laerd Statistics

Quiz Questions

1. What is the goal of an experiment?

- Maximize between group differences (treatment effect)

- Minimize error

- Statistic = between group s2/error

2. When is it appropriate to use a t test?

- When you have 1 IV

- When your IV has two levels (one per group)

3. What are the advantages of independent groups (random assignment)?

- Controls for unknown extraneous variables

4. What are the advantages of dependent (or correlated) groups?

- Controls better for extraneous variables and thus reduces error variance

- Results in a more powerful test

10.3 Interpretation: Making Sense of Your Statistics (See notes for example)

Interpreting Computer Statistical Output

• Understanding your statistics (What does all this mean?)

– What’s your research question? – What’s your hypothesis?

– What did you find?

• Communicating your findings in words (Using APA format to report results)

Quiz Questions

5. What important question should you keep in mind when analyzing statistical output?

- The research question: What is the effect of the IV upon the DV?

6. How should you write the hypothesis?

- Make sure you can infer an “if/then” relationship between the IV and the DV, which can be confirmed or disproven:

- “If the IV is manipulated, then the DV will be affected.”

Quiz Questions

7. What statistics should you inspect first, and what are you looking for?

- When interpreting output, inspect the descriptive statistics first to see the patterns among the means (i.e., DV scores) amongst groups (i.e., IV levels), the numbers of participants in each group (i.e., how big was the sample?), the standard deviations (i.e., how much the scores varied around the mean), and the standard error of the means (i.e., error).

8. What statistics should you inspect next, and what are you looking for?

- After you’ve inspected the descriptive statistics, look at the t test results.

- If you conducted an Independent measures t test, check Levene’s statistic. You want an insignificant result, meaning you’ve met the assumption of equal variances across groups. If not, use the stricter test (with fewer degrees of freedom).

- Next, examine the mean difference (i.e., difference between group means/DV scores), standard error of the mean (i.e., error term), upper and lower limits of the confidence interval (i.e., effect size estimate), t statistic (i.e., between group s2/error), df (i.e., n-1 + n-1), and p value (i.e., whether or not you’ve produced a statistically significant result).

Quiz Questions

9. How could the dependent samples t test be statistically significant, but the independent samples t test was not?

- The error term in the dependent samples t test was smaller, making the t statistic bigger and more likely to reach significance.

10. What statistics belong in the APA write up of your results?

- The means and standard deviations of each group and which were higher, the t statistic, the degrees of freedom, and the p value and the upper and lower limits of the confidence interval.

“The students exposed to treatment 1, control, reported higher anxiety scores (M = 26, SD = 7.54) than when exposed to treatment 2, talk therapy, (M = 21, SD = 6.56). The mean difference was significant, t(2) = 4.35, p = .049, CI [0.32, 9.97].”

10.4 The Continuing Research Problem

The Continuing Research Problem

•Research is a continuous, cyclical process

1. Literature review

2. Selection of IV

3. Two levels of IV

4. Random assignment OR repeated measures

5. Conclusions

6. New questions

7. Literature review…