PSY309 Lecture 2: Ethics and Measurement

What is the Tuskegee Syphilis Study? What are its major ethical violations?

1. Participants were not treated respectfully

2. The participants were harmed

3. The participants were a targeted, disadvantaged social group.

• Assumed trust, illiterate, poor disadvantaged individuals

• Waste away, resulted in death for many of participants

• Withheld information about treatment

• Intended benefit all people but conducted on disadvantaged group

The Milgram Obedience Study

• Ethical questions

• Balancing participant risk with the possible benefits that the research will have for society

• How authority influences responses, with something unethical

• Has to be balanced, how much risk

• Deception, participants deceived, not debriefed, forceful probes, leading participants to believe they could not drop out of the study.

Core Ethical Principles

1. Belmont Report

2. APA

What are the Belmont Report Principles and Applications? (3)

1. Principle of respect for persons

• Informed consent

2. Principle of Beneficence

• Assess potential harm to participants & potential benefits to society

3. Principle of justice

• Who bears the burden of research participation?

• Syphilis study, should be same group of people who should benefit from it.

What are APAs 5 General Principle?

• Belmont Plus 2

1. Beneficence

• Treat people in ways that benefit them

• Do not cause suffering

• Conduct research that will benefit society

2. Justice

• Strive to treat all groups of people fairly

• Sample research participants from the same populations that will benefit from the research

• Be aware of biases

3. Respect for people’s rights & dignity

• Recognize that people are autonomous agents. Protect people’s rights, including the right the privacy, the right to give consent for treatment on research & the right to have participation treated confidentialy.

• Understand some populations may be less able to give autonomous consent & take precautions against coercing such people.

4. Fidelity and Responsibility

• Establish relationships of trust

• Accept responsibility for professional behaviour (in research, teaching & clinical practice)

5. Integrity

• Strive to be accurate, truthful and honest in one’s role as researcher, teacher or practitioner

What are Ethical Standards for Research?

• Institutional review boards

  • Submit proposal before any research or doing anything

  • Monthly

  • Rounds of revision, modification & guide researchers

• Informed consent

  • Consent to be part of research study

  • Infants or very elderly people, defer to substitute decision maker, provide consent on behalf of dependant (infant, elderly person in long-term care)

• Deception

  • Essential for some experiments, inhibit researchers to discover hypothesis

  • Reserved for instances integral to research design

  • Deception shouldn't be involved in project not crucial

• Debriefing

  • Must debrief participants

  • Provide resources if cause anguish later on

  • Complete explanation of what went on

• Research misconduct

• Data fabrication & data falsification

• Plagiarism, self plagiarism

What are ethical standards for animal research?

• Legal protection for laboratory animals

• Animal care guidelines & the 3 R’s

  • Replacement, Refinement, Reduction

• Attitudes of scientists & students toward animal research

• Attitudes of animal rights groups

• Ethnically balancing animal welfare, animal rights & animal research

Ethical Decision Making: A Balance

• Requires a balance of priorities

• We need to weight the potential harm to human or animal participants against what the knowledge gained from the research will contribute to society/

Cost-Benefit Analysis

Facebook Violation of Ethical Principles

Institutional review board

• Was the study reviewed by an IRB?

• The study’s lead scientist was employed by Facebook and is not required to follow federal ethical guidelines such as the Common Rule

• The other 2 scientists had the study reviewed by Cornell University’s IRB. and decided the study did not fall under its program because the data had been collected by Facebook

• This example highlights that private businesses sometimes conduct research on people who use their products & such research might not be reviewed for ethics

Informed Consent

• Were Facebook users able to decide whether they wanted to participate?

• The study’s authors reported that when people create a Facebook account, they agree to a data use policy & this was constituted informed consent

• Not all critics agreed. Although it had agreed to public the paper, it was concerned that the study did not allow participants to opt out

Deception & Debriefing

• Were participants deceived about the study? If so, were they debrifed?

• Participants were not told their newsfeeds might have been manipulated for research purposes.

• Participants were deceived through omission of information

• People were not debriefed afterward. Even now, people cannot find out whether they participated in this study or not.

What would the IRB say about this study?

If an IRB has considered this study in advance, they would have evaluated it for…

1) Respect for persons

• Respect for persons requires informed consent. Participants did not consent to this study.

• Although people did not provide informed consent for this study, informed consent might not be deemed necessary when a study takes place in a public place where people can reasonably expect to be observed

2) Beneficence

• The study demonstrated that people felt worse when positive emotion posts were removed

• The researchers argued that the study benefited society. Social media plays a role in many people’s daily lives & emotions are linked to well-being.

• People may have suffered a bit but wa their distress any greater than it might have been in daily life? Perhaps not, because effect size was so small

• Facebook already manipulates newsfeeds, selecting stories & posts according to computerized algorithm.

• The results did not show that social media is a source of emotional contagion that could potentially affect public wellbeing.

3) Justice

• The study randomly selected hundred of thousands of people who read Facebook in English

• Because the sample was selected at random, it appears that the people who “bear the burden” of research participation were the same types who could benefit from its findings

• The principles of justice has probably been met.

What are Constructs?

• Constructs are Hypothetical Entities that explain observed phenomena

  • ex. culture, intelligence, empathy, companionship, stress

• Research questions are generally formed on constructs

  • Ex. Does intelligence predict productivity

• We generally cannot directly study or measure the constructs that we are interested in

• We use tools to measure or quantify the construct of interest

What does opperationalize mean?

• Opperationalize is the the process of turning abstract concepts or ideas into observable & measurable phenomena

Describing Variables

• When describing your variables, it is important to inform the reader about both the construct that is being investigated & how the construct is opperationalized

Measured vs. Manipulated Variables

• A Measured variable is observed & recorded

• A Manipulated variable is controlled

• Some variables can only be measured (ie. cannot be manipulated)

  • ex. gender, age, height cannot be manipulated

• Some variables can be both manipulated & measured, depending on the context/study

  • ex. anxiety can measure natural state levels of anxiety but can also do manipulation

Describing Variables Example Chart

Variable Opperationalization

Relationship Between Variable Types

• We run studies so that we can make claims about the variables that we are interested in studying

What are 3 types of claims?

1) Frequency Claims

• How often a given outcome occurs

2) Association Claims

• How levels on two or more variables might be related

3) Causal Claims

• X causes y

What are Frequency Claims?

• A frequency claim describes a particular level or degree of a single variable

• Frequency claims involve only one measured vairable

Example of Frequency Claim: Self-Reported Engagement

• No students indicated that they were “not at all engaged” at any point during class 1

• 80% of responses were 5 or higher on the engagement scale

• Students reported high levels of engagement in PSY309 on the surveys

• Negatively skewed distribution, more negatively skewed as class went on

Example of Frequency Claim: Level of Self Reported Anxiety

• 8% of responses indicates that students felt ‘not anxious at all’ about the Psy309 course content

• 8% of responses were ‘extremely anxious’ about the Psy309 course content

• Approximately normally distributed

Example of Frequency Claim: Self Reported Confusion

• 25% of responses were “not confused at all”

• 65% of responses were below 4 on confusion scale

• Students reported low levels of confusion & high levels of engagement with the Psy309 content

• More rightly skewed, positively skewed

What are Association Claims?

• An association claim argues that one level of a variable is likely to be associated with a particular level of another variable

• Association claims are supported by studies that have at least 2 measured variables

• Variables that are associated are said to correlate.

  

Example of Association Claim: Relationship between engagement & confusion

• Levels of self reported engagement do not correlate with levels of self reported confusion

Example of Association Claim: Anxiety and Confusion

• Higher levels of self reported anxiety are associated with higher levels of self reported confusion

Example of Association Claims: Self-Reported Anxiety & Levels of Engagement

• Self reported levels of anxiety & engagement do not correlatte

Making Predictions Based on Associations

• Some association claims are useful because they help us make predictions

• The stronger the association between the two variables, the more accurate the prediction will be

• Both positive & negative associations can help us make predictions, but zero associations cannot

What are Causal Claims?

• A causal claim argues that one variable causes changes in the level of another variable

• Causal claims are supported by experiments (studies that have manipulated variable & a measured variable)

Example of Causal Claim

Not all Causal Language is Supported by Research

• Not all claims we read about in the popular press are based on research

• Many (most) claims are based on experience, intuition or authority

What are the 4 Validities? Summary Chart

1. Construct validity

• How well the variables in a study are measured or manipulated/opperationalized

• The extent to which the operational variables in a study are a good approximation of the conceptual variables

2. External validity

• The extent to which the results of a study generalize to some larger population

3. Statistical validity

• How well the numbers support the claim, the extent to which the studys findings are reasonable & accurate

4. Internal Validity

• In a relationship between on variable A and another B, the extent to which A, rather than some other variable C, is responsible for changes in B

Interrogating Frequency Claims & the Big 4 Validities

Construct Validity

• How well a variable is operationalized

• Example: 39% of teens text while driving. How was this measured

External validity

• Generalizability

• Who was measured and who do they represent/no represent?

Statistical validity

• The extent to which the studys findings are reasonable & accurate

• Often involves looking the error of measurement to establish a confidence interval or margin of error around a point estimate

Interrogating Association Claims & the Big 4 Validities

Construct Validity

• What was measured? How to operationalize the construct?

External validity

• Does the association claim generalize to other populations, contexts, times or places

Statistical Validity

• Strength of associaiton

• Statistical significance

Interrogating Causal Claims

• 3 key criteria must be met to support use of causal claims

1) Covariance

• The study’s results show that as A changes, B changes

  • ex. high levels of A go with high levels of B, low levels of A go with low levels of B

2) Temporal precedence

• The study’s method ensures that A comes first in time, before B

3) Internal Validity

• The study’s method ensures that there are no plausible alternative explanations for the changes in B, A is the only thing that changed.

• Rules out other alternative/plausible explanations

• Most difficult to support.

Interrogating the 3 Types of Claims Using the Big 4 Validities