Psych 311 – Chapter 2: Understanding the Research Process

Psych 311 – Chapter 2

Understanding the Research Process Thorough & Easy-to-Understand Notes

Based on the uploaded Chapter 2 material. The notes preserve the chapter's concepts and terminology while explaining them in simpler language.

1. Theory versus Hypothesis

Theory: A theory is a broad, organized explanation of a phenomenon. It goes beyond simply describing observations by explaining how different variables, processes, and principles fit together. Easy memory: Theory = Big Explanation.

Hypothesis: A hypothesis is a specific prediction that comes from a theory. It tells researchers what they expect to observe if the theory is correct. Easy memory: Hypothesis = Specific Prediction.

Example: Bowlby's Attachment Theory: Attachment Theory explains how a child's relationship with a primary caregiver can influence emotional and social development. Observations such as distress during separation are organized into a broader explanation involving different attachment patterns.

2. Four Attachment Hypotheses

Secure Attachment: Moderate distress during separation, followed by relatively quick calming when the caregiver returns.

Avoidant Attachment: Less visible distress during separation and less interest in seeking comfort during reunion.

Anxious/Ambivalent Attachment: High distress during separation and difficulty calming down after the caregiver returns.

Disorganized Attachment: Inconsistent or unusual behavior, such as approaching the caregiver while simultaneously avoiding them.

3. How Theories Generate Hypotheses

Core relationship: if the theory is correct, then a particular outcome should be observed.

Process: Researchers ask what a theory implies and turn that implication into a testable hypothesis. Hypotheses can also focus on a component or process within a theory that has not yet been directly observed.

Competing theories: Especially useful hypotheses distinguish between competing theories by making different predictions.

4. Hypothetico-Deductive Method

Basic logic: Theory leads to Prediction leads to Hypothesis leads to Research or Test leads to Data leads to Conclusion.

Easy memory: Big Idea  Specific Prediction  Test it.

5. Characteristics of a Good Hypothesis

Testable and falsifiable: It can be tested scientifically, and there must be possible evidence that could disconfirm it.

Logical: It is informed by previous theories, observations, and logical reasoning—not a random guess.

Positive: It states an expected relationship or effect.

Direction of reasoning: Usually broad theory to deductive reasoning to specific hypothesis. When no theory exists, specific observations can contribute to inductive reasoning.

6. Variables

Definition: A variable is a quantity or quality that varies across people or situations.

Quantitative variable: Measured by assigning numbers, such as height, talkativeness, or depression level. Remember: Quantitative = Number.

Categorical variable: Measured using category labels, such as major, occupation, or nationality. Remember: Categorical = Category/Label.

7. Operational Definition

Definition: An operational definition precisely defines a variable in terms of how it will actually be measured.

Why it matters: Abstract psychological constructs such as depression cannot be directly observed like height. Researchers must transform them into observable/measurable indicators.

Example: Depression could be operationally defined using scores on a depression scale, number of depressive symptoms, or diagnostic status.

8. Population versus Sample

Population: The larger group about which the researcher wants to draw conclusions.

Sample: The smaller subset that actually participates in the study.

Representative sample: Researchers generally want the sample to resemble the population in important respects.

Easy memory: Population = Who You Want to Know about; Sample = Who You Actually Study.

9. Sampling Methods

Simple random sampling: Every population member has an equal chance of selection.

Convenience sampling: Participants are selected because they are nearby and willing. It is practical but may not represent the population well.

10. Experimental versus Non-Experimental Research

Experimental research: Used to test causal relationships. Researchers manipulate one or more variables and control extraneous variables.

Non-experimental research: Variables are measured as they naturally occur. It can describe characteristics and relationships and make predictions, but cannot establish that one variable causes another.

Key causal principle: Manipulation + control of alternative explanations allow experimental research to support causal conclusions.

11. Variables in an Experiment

Independent variable (4): The variable manipulated by the experimenter; the presumed cause. Memory trick: I.V = I change it.

Dependent variable (D.V): The variable measured by the experimenter; the presumed effect. Memory trick: D.V = Data/value I measure.

Extraneous variable: Any variable other than the dependent variable that could influence the outcome.

Confound: A specific extraneous variable that systematically varies along with the variable under investigation and therefore provides an alternative explanation for the result.

12. Laboratory versus Field Research

Laboratory study: Conducted in a laboratory. Usually offers greater control over extraneous variables and often higher internal validity, but may be less representative of real-world conditions.

Field study / field experiment: Conducted in a real-world, natural environment. Usually offers less environmental control and often higher external validity. A field experiment manipulates an I.V in a natural setting while controlling extraneous variables as much as possible.

13. Internal versus External Validity

Internal validity: The degree to which researchers can confidently infer a causal relationship between variables. Strong control of alternative explanations increases internal validity.

External validity: The degree to which findings can be generalized to other circumstances or settings, especially the real world.

Typical trade-off: Laboratory studies tend to favor internal validity; field studies tend to favor external validity. Field experiments can sometimes achieve both.

14. Descriptive Statistics

Purpose: Descriptive statistics organize and summarize a set of data.

Central tendency: Describes the typical/average/center of a distribution.

Dispersion: Describes how spread out the scores are.

Correlation: Describes the strength and direction of a relationship between two variables.

15. Mean, Median, Mode, and Dispersion

Mean: Average of a distribution.

Median: Midpoint of a distribution.

Mode: Most frequently occurring score.

Range / Standard deviation / Variance: Measures of dispersion. Variance is the standard deviation squared.

Easy memory: Mean = Average; Median = Middle; Mode = Most.

16. Correlation Coefficient

What it tells you: The strength and direction of the relationship between two variables.

-1.00: Strongest possible negative relationship.

0: No relationship.

+1.00: Strongest possible positive relationship.

Positive correlation: As one variable increases, the other tends to increase.

Negative correlation: As one variable increases, the other tends to decrease.

Important: Correlation describes a relationship; it does not by itself establish causation.

17. Inferential Statistics & Statistical Significance

Inferential statistics: Allow researchers to draw conclusions about a population based on sample data.

Purpose: Used to judge whether observed differences or relationships may reflect a real effect rather than random chance.

Statistical significance: A statistically significant effect is one considered unlikely to be due to random chance under the chosen criterion.

5% threshold: The chapter uses the conventional 5% threshold: results with less than a 5% chance of being due to random error are treated as statistically significant.

Important: Statistics are probabilistic and do not provide absolute certainty.

18. Type 1 versus Type 2 Error

Type 1 error — False positive: Concluding that an effect is statistically significant when there is no real effect in the population. Memory: False Alarm.

Type 2 error — Missed opportunity: Concluding that results are not statistically significant when there is a real effect in the population. Memory: Missed Real Effect.

Threshold trade-off: Making the significance threshold more stringent reduces Type 1 errors but increases the chance of Type 2 errors. Type 2 errors are also more likely with very small samples.

19. Drawing Scientific Conclusions

Support: A statistically significant result consistent with a hypothesis can support the theory that generated the hypothesis.

Weaken/refute: A disconfirmed hypothesis weakens a theory or indicates that it needs refinement, but does not automatically prove the theory wrong.

Scientific language: Scientists generally avoid the term 'scientific proof' and emphasize scientific evidence rather than certainty.

20. Why a Hypothesis Can Be Disconfirmed

Type 2 error: A failed prediction may reflect failure to detect a real effect.

Faulty research design: The 4 may not have been successfully manipulated or the D.V may not have been measured well.

Unmet assumption: An unstated assumption of the theory may not have been met.

Repeated disconfirmations: Researchers should improve designs, modify theories, or eventually abandon theories that cannot account for the evidence.

21. Reporting the Results

Peer-reviewed journals: Results are typically reported in peer-reviewed journal articles. Psychology manuscripts typically follow A.P.A style.

Scientific conferences: Findings may be presented orally or as posters. Conferences provide opportunities for feedback before more rigorous journal peer review.

Book chapters: Book chapters can also report findings, preferably after editorial peer review.

22. Complete Research Process

1. Find a topic/research idea: Start with something worth investigating.

2. Review the research literature: Learn what previous research has found.

3. Generate and refine a research question: Turn a general topic into a specific question.

4. Evaluate interestingness and feasibility: Ask whether the answer is in doubt, fills a literature gap, and has practical implications; also consider whether the study is feasible.

5. Develop a theory-based hypothesis: Create a specific prediction based on theory.

6. Define variables operationally: Specify exactly how variables will be measured.

7. Select population, sample, and design: Decide who the study is about, who participates, and how it will be conducted.

8. Conduct the empirical study: Collect observations/data.

9. Analyze the data: Use descriptive and inferential statistics.

10. Draw evidence-based conclusions: Determine what the evidence suggests about the hypothesis and theory.

11. Report and publish results: Communicate the findings.

12. Use the new literature to generate the next question: Research is cyclical; new findings can create new research questions.

23. High-Priority Exam Review

Theory versus hypothesis: Theory = broad explanation; hypothesis = specific testable prediction.

I.V versus D.V: I.V = manipulated/presumed cause; D.V = measured/presumed effect.

Experimental versus non-experimental: Experimental research uses manipulation and control to examine causal relationships; non-experimental research observes variables naturally and cannot establish causation.

Internal versus external validity: Internal = confidence in causal inference; external = generalizability.

Type 1 versus Type 2: Type I = false positive; Type 2 = missed real effect.

Population versus sample: Population = entire group of interest; sample = smaller group actually studied.

Mean/median/mode: Mean = average; median = middle; mode = most frequent.

Research cycle: Question arrow hypothesis arrow operational definitions arrow design slash sample arrow study arrow data analysis arrow conclusion arrow report arrow new question.

Quick Memorization Sheet

Theory = Big explanation

Hypothesis = Specific prediction

Operational Definition = Exactly how a variable is measured

Population = Entire group of interest

Sample = Smaller participating group

I.V = Manipulated / presumed cause

D.V = Measured / presumed effect

Confound = Alternative explanation

Internal Validity = Confidence in causation

External Validity = Generalizability

Mean = Average

Median = Middle

Mode = Most frequent

Correlation = Strength + direction of relationship

Type 1 Error = False positive

Type 2 Error = Missed real effect