Psychology 272: Research Methods and Data Analysis II Review

Sources of Knowledge and the Scientific Method

• Sources of Knowledge: Research and understanding are built upon various foundations of knowledge acquisition:

  • Superstition: Knowledge based on subjective feelings, belief in magical events, or interpreting random events as omens.

  • Intuition: Knowledge gained without conscious reasoning; a "gut feeling."

  • Authority: Knowledge accepted because it comes from a respected or influential source.

  • Tenacity: Knowledge held onto simply because it has been believed for a long time, despite potential evidence to the contrary.

  • Rationalism: Knowledge derived through logical reasoning and the application of rules of logic.

  • Empiricism: Knowledge gained through direct observation and experience using the senses.

  • Science: A systematic method of combining rationalism and empiricism to test ideas against reality.

• Popper’s Principle of Falsifiability: For a theory to be considered scientific, it must be stated in such a way that it can potentially be proven false. Scientific claims must be testable and refutable.

• Research Methods:

  • Descriptive Methods: Used to describe behaviors and environments without manipulating variables.

  • Predictive Methods: Used to identify relationships between variables to predict future outcomes.

  • Explanatory Methods: Used to determine cause-and-effect relationships.

• Correlation vs. Causation: Correlation does not equal causation. Determining a relationship between two variables does not mean one causes the other. Violation of this principle is a critical error in research interpretation.

• Experimental Design Components:

  • Independent Variable: The factor that is manipulated or controlled by the researcher.

  • Dependent Variable: The factor that is measured; it is expected to change in response to the independent variable.

  • Control Group: A group that does not receive the experimental treatment and serves as a baseline.

  • Placebo Group: A type of control group that receives a fake treatment (like a sugar pill) to control for participant expectations.

  • Experimental Group: The group that receives the manipulation or treatment being studied.

  • Comparison Group: A group used to compare the effects of different levels of the independent variable.

  • Random Assignment: The process of assigning participants to groups such that every participant has an equal chance of being in any group, which helps ensure groups are equivalent at the start of the study.

Literature Reviews and Ethical Standards

• Sources of Literature: Researchers utilize literature reviews, reference sources, and primary journal articles to build their knowledge base.

• Structure of a Peer-Reviewed Journal Article:

  • Abstract: A brief summary of the entire study.

  • Literature Review: A summary of previous research on the topic and the rationale for the current study.

  • Methodology: A detailed description of how the study was conducted, including Sample, materials, and procedures.

  • Results: The statistical findings of the study.

  • Discussion: An interpretation of the results, their implications, and limitations.

• Institutional Review Boards (IRBs): Committees that review research proposals to ensure they meet ethical standards. The process can be rigorous; it is jokingly suggested that some IRBs might approve a study only in exchange for one's first born child.

• Ethical Requirements:

  • Informed Consent: Participants must be told about the nature of the study and give their voluntary agreement to participate.

  • Risk and Deception: Researchers must minimize risks to participants and justify any use of deception, ensuring participants are debriefed afterward.

  • Vulnerable Populations: Special care and ethical considerations must be applied when conducting research with children or animals.

Variables and Measurement Properties

• Defining Variables:

  • Operational Definitions: Defining a variable in terms of the specific procedures or operations used to measure or manipulate it.

• Types of Data:

  • Discrete vs. Continuous: Discrete data consist of separate, indivisible categories (e.g., number of children). Continuous data can take on any value within a range (e.g., time, height).

  • Categorical vs. Quantitative: Categorical data represent qualitative labels (e.g., gender, eye color). Quantitative data represent numerical amounts (e.g., weight, temperature).

• Properties of Measurement:

  • Identity: Different values represent different categories.

  • Magnitude: Values can be ranked in order of size or importance.

  • Equal Unit Size: The distance between units on the scale is the same throughout (e.g., the difference between $1$ and $2$ is the same as between $10$ and $11$).

  • Absolute Zero Point: A value of zero represents the complete absence of the variable being measured.

• Scales of Measurement:

  • Nominal: Only has the property of identity (labels or names).

  • Ordinal: Has identity and magnitude (ranking).

  • Interval: Has identity, magnitude, and equal unit size (no absolute zero; e.g., Fahrenheit temperature).

  • Ratio: Has all four properties: identity, magnitude, equal unit size, and an absolute zero point (e.g., weight, height).

Measurement Reliability and Validity

• Types of Measures: Self-report, tests/assessments, behavioral measures, physical measures, and cognitive measures.

• Reliability: The consistency or stability of a measure.

  • Measurement Error: The difference between the true score and the observed score.

  • Method Error: Errors resulting from the testing situation or the researcher.

  • Trial Error: Errors resulting from inconsistencies in the participant.

• Validity: The extent to which a measure actually measures what it claims to measure.

  • Face Validity: Whether the test looks like it measures what it's supposed to.

  • Criterion Validity: The extent to which a measure relates to a specific outcome.

  • Concurrent Validity: The relationship between a new measure and an existing established measure taken at the same time.

  • Predictive Validity: The ability of a measure to predict future behavior.

  • Content Validity: Whether the items on the test represent the entire range of the material intended to be covered.

  • Construct Validity: The extent to which a test measures the theoretical construct it is intended to measure.

  • Consequential Validity: The social consequences of using a particular test.

• Relationship Between Reliability and Validity: A measure can be reliable without being valid, but it cannot be valid unless it is reliable.

Observational, Qualitative, and Survey Methods

• Observational Methods:

  • Naturalistic vs. Laboratory Observation: Observing behavior in a real-world setting versus a controlled environment.

  • Expectancy Effects / Confirmation Bias: When a researcher's expectations influence the results of the study or their interpretation of the data.

  • Narrative Records: Detailed descriptions of behaviors as they occur.

  • Data Reduction Techniques: Converting qualitative narrative records into quantitative data for analysis.

• Other Research Methods:

  • Case Study Methods: In-depth investigation of a single individual or small group.

  • Archival Methods: Using existing records or data sets to conduct research.

  • Qualitative Methods: Includes Interpretivism, social construction, and Grounded Theory.

• Survey Methods:

  • Construction: Designing effective questions.

  • Administration: Methods include mail, phone, and electronic surveys.

  • Socially-Desirable Responding: The tendency of participants to answer in a way that makes them look good rather than being honest.

• Sampling:

  • Probability Sampling: Every member of the population has a known chance of being selected. Types include random, stratified random, and cluster sampling.

  • Nonprobability Sampling: The selection of participants is not random. Types include convenience (using whoever is available) and quota sampling.

Organizing and Analyzing Data

• Visualizing Data: Frequency distributions, bar graphs, histograms, frequency polygons, and ogive curves.

• Measures of Central Tendency:

  • Mean: The mathematical average.

  • Median: The middle score in a distribution when ranked.

  • Mode: The most frequently occurring score.

• Measures of Variability:

  • Range: The distance between the highest and lowest scores.

  • Average Deviation: The average extent to which scores differ from the mean.

  • Standard Deviation: A measure of the average distance of scores from the mean.

• Distributions:

  • Normal Distributions: Symmetrical, bell-shaped curves.

  • Skewed Distributions: Distributions where scores pile up on one end.

  • Kurtosis: The peakedness or flatness of a distribution.

• z-Scores: These allow for the comparison of scores from different distributions by standardizing them. They represent the number of standard deviations a score is from the mean in a standard normal distribution.

Correlations and Linear Regression

• Correlations: Measures the relationship between two variables.

  • Magnitude and Strength: Indicated by the correlation coefficient, ranging from $-1.00$ to $+1.00$.

  • Positive vs. Inverse (Negative) Correlations: In positive correlations, variables move in the same direction. In inverse correlations, they move in opposite directions.

  • Scatterplots: Graphical representations of the relationship between two variables.

  • Coefficient of Determination: Calculated as $r^2$, representing the proportion of variance in one variable that can be predicted from the other.

• Potential Problems in Correlation:

  • Intervening Variables: Unmeasured variables that might explain the relationship.

  • Curvilinear Relationships: Relationships that are not linear.

  • Restriction of Range: When the data set does not include the full range of possible values.

  • Heterogeneous Subsamples: When the sample consists of distinct groups that might mask the true relationship.

  • Outliers: Extreme scores that disproportionately affect the correlation.

• Regression: Synonymous with prediction.

  • Variables and Axes: Predictor variable on the $X$-axis and criterion variable on the $Y$-axis.

  • Regression Formula: $y = bx + a$

  • Slope: Represented by $b$.

  • Y-intercept: Represented by $a$.

Hypothesis Testing and Statistical Errors

• Hypotheses:

  • Null Hypothesis ($H_0$): The statement that there is no effect or no difference. This is the only hypothesis that is statistically tested.

  • Alternate Hypothesis ($H_1$): The statement that there is an effect or a difference.

  • If the "P" value is low, then the $H_0$ must go: We reject the null hypothesis if the probability of the results occurring by chance is very small.

• Test Types:

  • Directional (One-tailed): Specifies the direction of the expected effect.

  • Nondirectional (Two-tailed): Tests for any difference, regardless of direction.

• The Four Steps of Hypothesis Testing:

  1. Make a statement about the world ($H_0$).

  2. Test that statement.

  3. Make the appropriate conclusion about $H_0$ (Reject or Fail to Reject).

  4. Make the appropriate decision based on your conclusion.

• Decision Errors:

  • Type I Error ($\\alpha$): Rejecting the null hypothesis when it is actually true (False Positive). Described as a "sin of commission," "trippin'," or convicting an innocent man. Relationship example: rejecting a good partner mistakenly.

  • Type II Error ($\\beta$): Failing to reject the null hypothesis when it is actually false (False Negative). Described as a "sin of omission," "cluelessness," or letting a guilty man go free. Relationship example: being the "best you never had" because you were passed over.

  • The Smoke Detector Example: A Type I error is a smoke detector going off when there is no fire (annoying, false alarm). A Type II error is a smoke detector remaining silent when there actually is a fire (dangerous, missed signal).

• Specific Statistical Tests:

  • Single-group Design: Testing one sample against a known population.

  • z-test: Used when the population mean and population standard deviation are known.

  • One-sample t-test: Used when the population standard deviation is unknown and must be estimated from the sample. Degrees of freedom for this test are calculated as $df = N - 1$.