Analytic statement: statement that is always true.

Contradictory statement: statement that is always false.

Deductive model: reasoning from general principles to specific predictions. The deductive approach is used to test the assumptions of a theory.

Discussion: concluding section of the research report, used to integrate the experimental findings into the existing body of knowledge, showing how the current research advances knowledge, increases generalizability of known effects, or contradicts past findings.

Experimental hypothesis: statement that predicts the effect of an IV on the DV.

Falsifiable statement: statement that can be contradicted by experimental evidence.

Fruitful statement: statement that generates new studies. This is the heuristic function of an experimental hypothesis.

Hypothesis: prediction of a relationship between two or more variables.

Inductive model: reasoning from specific cases to general principles to form a hypothesis.

Introduction: beginning section of a research report that guides the reader toward your research hypothesis; includes a selective review of relevant, recent research.

Intuition: unconscious problem-solving.

Meta-analysis: a statistical reviewing procedure that uses data from many similar studies to summarize and quantify research findings about individual topics.

Nonexperimental hypothesis: prediction of how variables might be correlated, but not causally related.

Parsimonious statement: statement with a minimum number of required assumptions.

Psychological journal: a periodical that publishes individual research reports and integrative research reviews, which are up-to-date summaries of what is known about a specific topic.

Serendipity: finding things not sought.

Synthetic statement: statement that can be either true or false. Experimental hypotheses must be synthetic statements.

Testable statements: a statement that can be assessed by manipulating an IV and measuring the results on the DV.

Group 1: (a) What is a hypothesis? (b) Differentiate a nonexperimental hypothesis from an experimental hypothesis.

Group 2: (a) Why must a hypothesis be a synthetic statement? (b) How does this relate to falsification? (c) Why can't an experimental hypothesis be analytic or contradictory? (d) Why is testability important?

Group 3: (a) Why should research hypotheses be parsimonious? (b) Provide an example of a fruitful hypothesis.

Group 4: (a) Explain induction. (b) How did Skinner illustrate the inductive model?

Group 5: (a) Explain deduction. (b) How did Walster and colleagues illustrate this approach?

Group 1: How can researchers combine induction and deduction?

Group 2: (a) How could researchers build on the Bechtol and Williams (1977) study of beach litter? (b) What is the value of prior research to experimenters?

Group 3: (a) What is intuition? (b) What role does intuition play in research?

Group 4: (a) How did Pavlov illustrate serendipity? (b) Why is a meta-analysis a valuable source of information?

Group 5: What is the purpose of the Introduction section of an APA paper?

Experimental research depends on developing hypotheses. The first part of the chapter examines nonexperimental and experimental hypotheses. A nonexperimental hypothesis is not a statement about cause and effect. Instead, it predicts the relationship a researcher expects to find between two or more variables. An experimental hypothesis provides a tentative causal explanation of an event or behavior. It must be a synthetic statement that is falsifiable and testable. It should also be parsimonious, which means it requires the fewest supporting assumptions. The most fruitful or heuristic hypotheses stimulate further research. The authors explain that experimental hypotheses cannot be analytic or contradictory statements because these statements cannot be falsified. Since analytic statements are always true and contradictory statements are always false, we do not need to conduct experiments to test them.

The second part of the chapter examines complementary sources of experimental hypotheses. Researchers develop experimental hypotheses using induction, deduction, prior research, serendipity, intuition, and observation. The authors examine the processes of induction and deduction, and emphasize that researchers use both to generate testable hypotheses. Induction helps us devise general principles and theories that organize, explain, and predict behavior. Deduction allows us to test the implications of these principles and theories.

A review of prior research can help narrow down possible explanations and generate experimental hypotheses. Nonexperimental studies can suggest causal explanations that can be tested by experimental hypotheses. Serendipity, the knack of finding something you did not seek, can be a valuable source of hypotheses when we are open-minded and can appreciate the significance of an unexpected observation. Intuition is most likely to generate valuable hypotheses when it comes from experts. Finally, everyday observation of behaviors and real-world problems can suggest hypotheses when all else fails.

The chapter concludes with strategies for conducting a literature search. The authors explain the value of a meta-analysis, describe the functions of the Introduction and Discussion sections of a research report, and review library aids (e.g., PsycINFO and PsycLIT) used to locate relevant journal articles.
 

Hypotheses

The experimental hypothesis predicts the causal relationship between an IV and the DV. These have the form of a synthetic statement which can be true or false. Example: anxiety (IV) increases affiliation (DV). A nonexperimental hypothesis predicts how events, traits, or behaviors might be related. It is not a statement about cause and effect. Example: college grades should be related to ACT scores.

Properties of an experimental hypothesis

Experimental hypotheses must be synthetic statements that are testable, falsifiable, parsimonious, and fruitful. Synthetic statements can be either true or false. ("If you look at an appealing photograph, then your pupils will dilate"). In contrast, analytic statements are always true ("I'm in Missouri or not in Missouri") and contradictory statements are always false ("I am a PSYC major and I am not a PSYC major").

Testable means that we can manipulate the antecedent conditions and measure resulting behavior. Falsifiable means that research findings must be able to contradict the hypothesis. Parsimony means that we prefer the simplest hypothesis that predicts the causal relationship between the IV and DV. An experimental hypothesis is fruitful when it leads to new studies.

Models

The inductive and deductive models help us develop hypotheses. The inductive model goes from specific examples to general rules. Skinner advocated this approach in developing principles of operant conditioning.

The deductive model goes from general rules to specific predictions. This is easiest when a theory has clearly stated premises. Walster, Walster, and Berscheid's equity theory is an example.

Mendeleyev combined both approaches. He moved from induction (finding a pattern of atomic weights) to deduction (predicting the correct weights of elements).

The research process

Search the experimental literature to develop hypotheses. The Introduction section of a psychological journal article traces the investigation of a research area and identifies a question that has not been definitively answered. The Discussion section places the experimental findings in the context of your literature review. In this section, the researcher explains how the findings support or contradict prior reports.

As you observe behavior in the laboratory or field, take advantage of serendipity, which is finding things not sought. Pavlov's discovery of classical conditioning is an example of serendipity. These events are only useful when the researcher is prepared to grasp their significance.