CH 2 - Methods in psychological science

Empiricism?

Empiricism: The conviction that accurate knowledge of the world can be acquired by observing it

Scientific method: A method for using empirical evidence to establish facts

Fact?

an objective statement, usually based on direct observation, that reasonable observers agree is true

Theory?

a hypothetical account of how and why a phenomenon occurs

-explain things

Hypothesis?

a specific testable prediction made by a theory

What are the steps to the scientific method? Think of the acronym

IGTAB

1. Identify - identify question of interest

2. Gather - gather information and form hypothesis

3. Test - test hypothesis by conducting research

4. Analyze - Analyze data, draw tentative conclusions, and report findings

5. Build - build body of knowledge and restart the process

Step 1: Identify a question of interest?

Questions ay come from current or historical events, observations of day to day experience, various forms of media, and scientific articles (alternative perspectives, expanding scope)

Step 2: Gather information and form hypothesis?

Look to see what work has been done and get an idea of relevant processes. Form a hypothesis, a prediction of what would happen if your ideas are true

Step 3: Test hypothesis by conducting research?

Apply your ideas/hypothesis to a research design and collect data

Step 4: Analyze data, draw tentative conclusions, and report findings?

Identify the proper methods to analyze data, makes of the results and share. Report in journals, conferences, books, etc.

Step 5: Build a body of knowledge?

Follow-up with new hypotheses, applications, connections, etc. Combine ideas to make an overall theory, a set of formal statements that explains how and why certain events are related to one another. You may need to refine, replace, or abandon theories

Why should the sicentifi method be used?

Helps reduce various biases

-hindsight bias

Help to provide better theories

-prediction and control

More objective than subjective

What makes a good theory?

1. Comprehensive - it covers/addresses previous knowledge/observations

2. Testable/Falsifiable - can be proven wrong - it can be tested

3. Simple Law of parsimony: if two theories can explain and predict the same phenomena equally well, the simpler theory is preferred

-Less assumptions = better

4. Generative

-new ideas/insights

-New predictions

-it generates new ideas, applications, insights, and hypotheses

Variable is?

-a conceptual definition vs operational definition

-concept - abstract understanding

-operational - how we observe ina s tudy

ex - people score on a self-esteem scale

Variable: A property that can have more than one value

What is a self-report measure? What's the main challenge and advantages?

Self report measures: those that ask people to report their own knowledge, beliefs, experiences, or behaviours

-ex - surveys

Easy to use but requires caution: people may be unwilling or incapable of providing honest or accurate responses

-ex - social desirability bias

WHat is a behavioural observation measure? Whats the main challenge or advantages?

Psychologists may use the occurence, frequency, or timing of a behavioural occruence

Knowing one is being observed may affect behavior

-habituation

+accustomed to something and be one self

-unobtrusive measures

-archival measures

Behavioural observation - Habituation?

-accustomed to something and be one self

-accustomed to their conditions

Habituation: A general process in which repeated or prolonged exposure to a stimulus results in a gradual reduction in responding

Behavioural observation - unobtrusive measures?

Observing

Unobtrusive measures: A measure obtained without disturbing the participant or alerting them that a measurement is being made

Behavioural observation - archival measures?

Past data "archive"

Archival measures: Data obtained from existing records and storage such as past research, newspapers, government records and documents.

Physiological measures? Types?

(physical measures):

Types

-heart rate

-sweat

-neuroimaging

Hard to fake these measures but also can be hard to interpret

Becoming very important in most areas in psychology

Physiological measure?

Physiological measure: any of a set of instruments that convey precise information about an individual's bodily functions, such as heart rate, skin conductance, skin temperature, cortisol level, palmar sweat, and eye tracking.

What are the two types of methodologies of research designs?

1. Qualitative research

2. Quantitative research

Case Study? Purpose?

Case Study: An in-depth analysis of an individual, a group, or an event

Data may be collected in various ways

Purpose: Discover principles that can generalize to other people or events

-However, they may not

Advantages of a case study?

In-depth analysis

May challenge previously held beliefs or theories

May be generative

What are the disadvantages of a case study?

Cannot determine causation

May not generalize

Potential bias

Naturalistic Observation?

Naturalistic observation: Research where behavior is observed as it occurs in a natural setting

-The use of coding systems allow for systematic measuring

What are the advantages of a naturalistic observation?

Advantages: Observing behavior in real (vs. artificial) settings and can provide a wealth of information

What are the disadvantages of a naturalistic observation?

Disadvantages: Cannot determine causation, generalizations are limited, and the presence of an observer may influence behavior

Survey Research?

Survey research: Research involving the collecting of information about a topic by administering questionnaires to or interviewing people

Note: Surveys (as a tool) can be used in non-survey research as well!

What are the advantages of a survey?

Advantage: Efficient to collect and can be informative

What are the disadvantages of a survey?

Disadvantages: No cause and effect, self-reports can be inaccurate, unrepresentative samples can lead to invalid conclusions, and inferences are based on probabilities

Correlational Research?

Correlational research: A research design in which the researcher examines the relationship between two (or more) variables

-Strictly Quantitative

-May include groups/categories but are represented numerically (e.g., Dummy-coding)

Data may come from various methods

How does correlational research allow us to make predictions?

Allows us to make predictions

-Predictor (x) and criterion/outcome (y)

-X - independent variable

-Y - dependent variable

Correlation coefficient?

Correlation coefficient: A statistic that indicates the direction and strength of the relation between two variables

-Range is from -1.00 to +1.00

How to measure direction of correl?

Direction is indicated by whether it is + or -

-Positive -> As one variable increases, so does the other

-Negative -> As one variable increases, the other decreases

How to measure strength of correl?

Strength is indicated by the number

-Closer to -1 or +1 is stronger; 0 represents no relationship

What do we always have to recall about correl?

Recall that correlation does not imply causation

Correlation Does Not Imply Causation?

The third variable problem: X will seem to cause Y and have a good correl but there may be no causal relationship between X and Y in reality. A third variable may be the causal factor

Correlational Research advantages?

Allows us to see if there is a relationship and make predictions

Establish a relationship to study under more controlled conditions

Study phenomena that are unethical or impractical to study in the lab

Correlational Research disadvantages?

Cannot determine causation

-X causes Y

What is special about Experiments?

Unlike the other designs, experiments can determine causation

Experiments have three essential characteristics?

1. The researcher one (or more) variables

manipulates

2. The researcher whether this manipulation influences other variables

3. measures

The researcher attempts to extraneous factors

control

Independent variable (IV)?

The factor that is manipulated or controlled by the experimenter

-Causality

+manipulation

Dependent variable (DV)?

The factor that is measured by the experimenter and that may be influenced by the independent variable

-Effect

+measure

3 criteria for Causation?

1. Covariance

2. Temporal precedence

3. Internal validity

Temporal precedence?

Temporal precedence: The manipulation (IV) occurs before the measurement (DV)

Internal validity?

Internal validity: The degree to which an experiment supports clear causal connections

-Is the study doing what it is intended to do and nothing more?'

-Is it doing what its supposed to in an accurate way

Experiment - Experimental (treatment) group?

Experimental (treatment) group: A group that receives a treatment or an active level of the independent variable

Experiment - Control group?

Control group: A group not exposed to the treatment or receives a zero-level of the independent variable

-Used as a comparison group - the baseline

Repeated measures (experimental design)(or within subjects) design - Counterbalancing? What does it eliminate?

Counterbalancing: A procedure in which the order of conditions is varied so that no condition has an overall advantage relative to the others

Eliminates several alternative explanations:

-Practice effect (performance will get better)

-Fatigue effect (performance will get worse)

-Spillover effect (effect of one condition will 'spillover' to another)

-Habituation (conditions will lose their effect)

Experiments with Multiple Independent Variables - It is common for experiments to have more than one independent variable

It is common for experiments to have more than one independent variable:

-Attempts to address the complexity of everyday life

-Often 2 or 3, any more gets too complicated

-May combine between- and within-subjects IVs in a single study

What are the 4 disadvantages of experiments?

1. Confounding variables

2. May not generalize to "real-life"

3. Placebo effects

4. Experiemnter/observer expectancies

Confounding variables?

-things may be confused with the independent variable

-COD for example has many causes for aggression other than just being a violent game as compared to animal crossing

Confounding variable: In an experiment, an unaccounted variable that is conceptually distinct but empirically separate from one or more other independent variables. Confounding makes it impossible to differentiate that variable's effects in isolation from its effects in conjunction with other variables.

Placebo effects?

being in a study may produce the effect - power of the mind

Experimenter/observer expectancies?

-may subconciously or non that causes things

-may cause inaccuracies

-DEMAND CHARACTERISTICS

What are the criteria of quality research?

-Does the sample reflect the population?

+the participants we choose out of the population

-Reliability

-Validity

-Generalizability

Reliability?

Reliability: The tendency for a measure to produce the same result whenever it is used to measure the same thing

What are the two criteria of reliability?

1. If used more than once, would results be similar?

-consistency in measurements

-weighing myself

2. Measurement error, not bias

-answering questions to look better - bias

What are the three types of reliability?

1. Test-Retest reliability

2. Interrater Reliability

3. Internal reliability

Test retest reliability?

Test-Retest reliability: When the assessment is consistent at different points in time

-score almost the same all the time

Interrater reliability?

Interrater Reliability: When assessments from different researchers/evaluators are consistent

-score the same from different evaluators

Internal reliability?

Internal reliability: When items within an assessment are consistent

-same tests

-same knowledge

-same criteria

What are the three types of reliability basically?

Basically consistency in everything

Validity?

Validity: The characteristic of an observation that allows one to draw accurate inferences from it

-Does it measure/predict what it intends to? +Conceptual definition -> Operational definition?

What is validity basically?

Basically accuracy

What are the 4 types of validity?

1. Face validity:

2. Content validity:

3. Convergent validity:

4. Discriminant validity:

Face validity?

Face validity: When the measure to assess what it is meant to measure appears

Content validity?

Content validity: When the measure assesses all of the parts of a defined construct a

-midterm only measures what chapters it says

Convergent validity?

Convergent validity: When the measure correlates more strongly with similar/same constructsz

-hitting a punching bag causes higher aggression

-is it predicting what it should be (correlated) prediction through the relationship

Discriminant validity?

Discriminant validity: When the measure correlates weakly or not at all with distinct constructs

-does it measure with precision

Reliability and validity?

Grouping = Reliability

Bullseye = Validity

What are the 4 types of reliability and validity when it comes to getting a bullseye?

(A) Neither valid nor reliable

-not consistent or valid

(B) Reliability but not valid.

-consistent but not valid

(C) Reliable and valid

(D) Not reliable but valid can hit the bullseye -but not consistently in the same spot

Generalizability?

The process of deriving a concept, judgment, principle, or theory from a limited number of specific cases and applying it more widely, often to an entire class of objects, events, or people

Threats to generalizability?

Undergraduate samples, WEIRD samples, gender bias, cohort effects

Cohort effects?

Cohort effects: Outcomes associated with a member of a group whose members all undergo similar experiences. May be different from age and period effects in research, as well as overlap (those of a similar age or from a particular period of time are likely to have more similar experiences than others)

Two Types of Statistics?

1. Descriptive statistics

2. Inferential statistics

Descriptive statistics?

Descriptive statistics: Summarize and describe characteristics of a data set (also called a distribution)

-E.g., Average course grade in the class

Inferential statistics?

Inferential statistics: Tells us how confident we can be in drawing conclusions or inferences about a population based on findings obtained from a sample

-E.g., Does your grade in this class predict outcomes in other psychology courses?

-trying to make an inference on what information we have

Central Tendency?

Measures of central tendency: Measures that describe a distribution in terms of a single statistic that is in some way "typical" of the sample as a whole

Mode?

Mode: The most frequent response in a distribution

Mean?

Mean: The mathematical average of a set of scores

-Denoted as M in journal articles (e.g., M = 4.5)

Median?

Median: When ordered from lowest value to highest value, the score that is in the middle of the distribution

-If there are an even number of observations, the median will be the average of the two middle values

Means are usually used to?

Means are usually used to describe central tendency (and compare groups for inferential statistics)

Variability?

Measures of variability: Measures that provide information about the spread of scores in a distribution

empirical rule?

Standard deviations from the mean percentages: +/-1=34.1% on each side of the mean (68.2% total); +/-2=13.6* (95.4%); +/-3=2.1% (99.6%);

Range?

Range: The difference between the highest and lowest score in a distribution

-May not be that informative

Standard deviation?

Standard deviation: A common measure of variation around the mean

-The square root of the variance and denoted as SD in journal articles and reported alongside the mean (e.g., M = 4.5, SD = 0.98)

Statistical significance?

Statistical significance: A term that suggest that it is unlikely that a particular finding occurred by chance alone

pvalue?

pvalue: The probability of finding the observed, or more extreme, results when there is no true effect

Psychology mostly uses p

Psychology mostly uses p < .05 (⍺ = .05)

Effect size?

Effect size: How large of an effect there is or how strong of a relationship is there

Effect size: May be discussed in context of effect in relation to standard deviations; correlation coefficient is also a measure of effect size

E.g., Test a weight-loss drug 1,000 people taking the drug (treatment) vs. 1,000 not (control) Statistical difference even if averages differ by 1 kg; Is it worth it? How about if the standard deviation is 1 SD?

Psychology mostly uses p < .05 (⍺ = .05) ?

Psychology mostly uses p < .05 (⍺ = .05)

-0.05 and above means no effect or result - accept the null hypothesis

-P < 0.05 - reject the null hypothesis