Lecture 2: Research Method and Ethics.

This is based on the Slides of Lecture Two from professor Glen Schafe at CUNY Hunter College. Includes 1. Research and Data Methods 2. Data and Analysis Reporting 3. Research Ethics.

What are the five ways of "knowing things"?

1. Intuition
2. Deference to Authority
3. Rational Tought
4. Putting your ideas to the to the test!

Intuition

Way of "knowing" Things.

Rational Thought

Based on logic and reasoning.

Putting your ideas to the test

Based on systematic observations, comparisons, data, collection and analysis.

Why is Empirical Investigation critical for science?

It grounds knowledge in observable, measurable, and testable real-world evidence rather than relying solely on theory or belief.

Step 1: Scientific Method

Use your theory to create a testable question.

Step 2: Scientific Method

Design a study to ask your question.

Step 3: Scientific Method

Perform your study

Step 4: Scientific Method

Analyze your findings, support, or modify your theory as necessary and ask new questions.

Theory

Is a set of interconnected ideas and statements used to explain a collection of observable facts.

Hypothesis

Is a testable predication (or question) derived from a theory about what happens at given circumstances.

What is the main purpose of Descriptive Research?

Observe and record (e.g. 'describe' behavior)

How is Descriptive Research conducted?

Case, studies, surveys, naturalistic observations.

What are the Weaknesses of Descriptive Research?

No control over the variables; may be misleading if too few observations are made.

What is the purpose of Correlational Research?

Detect naturally occurring relationships; asses how well one variable can be used to verdict another variable.

How is Correlational Research conducted?

Compute statistical association (a.k.a "correlation") either among existing data sets or survey responses.

What is being manipulated in Correlation Research?

Nothing!

What are some of the Weaknesses of Correlation Research?

No control over the variables; does not allow the researcher to draw conclusions about clause and effect relationships.

What is the research method of experimental research?

Explore "cause and effect relationships"!

How is experimental research conducted?

Deliberately manipulate one or more variables; use random assignment of participants to groups.

What is being manipulated in Experimental research?

The Independant Variables.

What are the weaknesses of Experimental Research?

Sometimes not feasible; results may not generalize to other contexts; not ethical to manipulate certain variables.

Descriptive Approaches Method 1:

Case Studies:
Examine one person (or social unit) in depth in the hope that the findings can be generalized to others.

Pros:
-Very detailed.
Cons:
-Time consuming
-And may be too narrow or be a bit general to use.

Descriptive Approaches Method 2:

Naturalistic Observations:
Observe and record behavior as it occurs in its natural settings.

Pros:
-It's real
Cons:
-But it may not reveal all the factors that contribute to the observed behavior.

Descriptive Approches Method 3:

Surveys:
Ask many people to respond in depth to a series of questions about their present or past behavior (or their experiences)

Pros:
-Relatively easy and provides many observations.
Cons:
-The participants may not respond truthfully to surveys or the sample may not represent of their larger population.

Random Sampling

Allows us make inferences about a whole population of people when only two are studied.

Method

Asks participants how often they exercise and how often they exercise and measure their mood using a survey. (x and y on a graph)

Correlation

Correlation measures the strength of the relationship between the two variables (e.g. how accurately we can predict the value of Variable X by knowing the Variable Y)

Positive Correlation

As the value of one variable increases the value of the other also increases.

Negative Correlation

As the value of one variable increases the value of the other decreases.

No Correlation

There is no linear relationship between the values of two variables.

The Correlation Coefficient (R)

Ranging from +I to -I, is a measure of how well the value of one variable can be predicted by knowing the value of the other.
(Allows us to predict and not to explain)

The Experimental Approach

1) Deliberately manipulate at least one variable in your hypothesized effect.

2) Precisely mesure something with a hypothesized relationship to the variable being manipulated.

Random Assignment

-The Key to every (good) experiment is random assignments of participants to groups.
-Random assignment of participants to groups.
-Random assignment spreads participant variability (individual differences) evenly between the groups.

Confound

A confound exists when there's a variable that is not "controlled for" (e.g. not randomly and evenly distributed between groups)

What does Random Sampling allows us to do?

Random Sampling allows use to make inferences about a whole population of people when only a few are studied.

Random Sampling

The first sample must be large enough and be selected at random for it to be representative of the entire population.

Random Assignment

The participants in the random sample are next randomly assigned to our conditions.

Experimental Variables

Independent and Dependent Variables

Independent Variable

The variable that is being deliberately manipulated in the experiment (eg. the variable that's created by the experimenter)

Dependant Variable

The variable that is ultimately measured in the experiment.

Aerobic Exercise

Physical Activity, typically prolonged and of moderate intensity (eg., walking, jogging, cycling) that involves the use of Oxygen in the muscles provided the needed energy.

Features of GOOD Experiments

-A 'control' condition:
-A group that does not receive the manipulation.

-'Blind' (or 'double-minded') participation.:
-Participants should not be aware of the specific hypothesis being tested until the end of the study because it could create demand characteristics.
-Researchers should (ideally) be unaware of which group participations belong to until the end of the study to until the end of the study to avoid experimenter demand.

-Replication
-Does the finding hold up in future experiments?

-External Validity:
-Does the finding relate to the 'real' world at all?

Why don't we always do experiments in psychology?

-Because it is not always feasible or ethical to do so!
-Psychology "big" questions often require correlational studies!

Descriptive Statistics

-There is always variability in data collection, but we can still define the central tendency.

Mean

The average score.

Median

The 50th percentile.

What is "Normal Distribution" on a graph

In a perfectly symmetrical normal distribution, the mean, median, and mode are all the same number!

Skewed Distrbution

An asymmetrical probability distribution where data clusters to one side, resulting in a long tail on either the left or right.

Variability

The standard deviation is a measure of variability in a sample (eg., how far scores spread out from the mean)
-Note that the mean (the average) is the same in each distribution!

Inferential Statistics

-Because variability always exists, we can never say with 100% certainty that our experimental manipulation was effective!
-We can, however, compute the probability that the difference has occurred by chance alone.
-This is called a test of statistical Significance.

Statistically Significant

If the probability of it being found due to CHANCE! is less than 5%.

-Statically significant does to mean "important"!

-It simply means that we are 95% confident that our finding (e.g. increased mood in exercise group) represents a real affect.

-Thats the best we can do!

Result (part of scientific method)

-Scientific findings in Psychology, like in all field of science, are submitted to journals for publication.

-Each journal is charged with rigorously reviewing articles prior towards publication.

-Most articles are peer reviewed.
1. Sound and Design
2. Impact

Method 1: How to prevent Ethical Science:
Ethical Review Boards:

Method 1: How to prevent Ethical Science:
Ethical Review Boards:

Method 2: How to prevent Ethical Science:
Informed Consent:

All participants must known what they will experience and must freely agree to participate.

Method 3: How to prevent Ethical Science:
Protection:

Of those who cannot give consent.
-Minors, prisoners, mentally surrounded individuals.

Your Rights as Participants:

-You may refuse to participant in any particular study.
-You may withward from a study while in progress without penalty (you will still receive credit)
-You may request that the data you provide be destroyed and not be used in any way.
-You may expect to be "debriefed" at the end of the study (e.g., have the study be explained to you)
-You may choose not to participate in research at all!
-An 'Alternative Assignment" will be made available.

Research with Animals

Research with non-humans vertebrate animals must also be reviewed in detail and pre-approved by the "Institutional Animal Care and Use Committee (IACUC)

-There must be no "reasonable alternatives" to the use of an animal model.
+Certian questions can only be asked using living, non-human animals (e.g. "Mechanisms of behavior X"?)

-Pain and/ or distress of animals used in research must always be mentored and minimized.
-This is both an ethical and a scientific justification (e.g., animals in pain or distress are unlikely to provide reliable data)