Experimental Psychology Chapter 3 & 4

Nonexperimental approaches - do not create levels of an independent variable nor randomly assign

subjects to these levels.

-         used where experiments are not ethical or possible, or where we want to test hypotheses in realistic conditions; used to study behaviors in natural settings to explore unique or rare occurrences.

Internal validity - the degree to which a researcher can establish a causal relationship between the independent and dependent variables.

-         an internally valid experiment allows us to draw cause-and-effect conclusions.

➔     An experiment has high internal validity when we can demonstrate that only the antecedent conditions are responsible for group differences in behavior.

➔     Laboratory experiments are often higher in internal validity because of their control of extraneous variables.

●       Researchers create levels of the IV and use procedures like matching and random assignment to conditions.

External validity - the degree to which research findings can be generalized to other settings and individuals.

➔     Nonexperimental studies achieve higher external validity because they are more frequently conducted in real-world settings with a more diverse sample of participants than experiments.

➔     What we gain in external validity we might lose in internal validity.

All approaches to research can be described along two major dimensions:

1.      Degree of manipulation of antecedent conditions - concerns assignment of subjects to

antecedent conditions created for the experiment.

➔     degree of manipulation in the treatments (IV) given; In the case of Nonexperimental designs, this is low.

➔     theoretically varies from low to high (from letting things happen as they will to setting up carefully controlled conditions).

2.      Degree of imposition of units - how much you limit a subject’s responses on the DV.

➔     Example: gathering data on teenage behavior (low imposition of units) versus gathering data on the music preferences of teenagers: e.g. “How often do you listen to hip-hop?” (high imposition of units).

-         involves a subject's description of personal subjective experience; Private Personal experience is the source of data

-         low in manipulation of antecedent conditions and low in imposition of units (low-low).

-         describes but cannot explain behavior

-         these days, it is seldom, if ever, used as a research method on its own; its elements are most often combined with other research methods.

-         a researcher compiles a descriptive study of a subject's experiences, observable behaviors, and archival records kept by an outside observer; heavily used in Forensic, Organizational, and Clinical psychology

-         range from low-low (low manipulation of antecedent conditions and low imposition of units) to low-high

Retrospective data - recollections of past events that are collected in the present.

➔     Example: childhood memories constitute retrospective data, but your undergraduate portfolio does not since it was collected in the past.

-         information may be compromised by faulty memory, current mood, and the retrieval cues that are present when you are asked to recall an event.

★      In field studies, the problem of reactivity can happen. This is when subjects alter their behavior when they know that they are being observed

○       Your baby sister stops saying “Gramma” when you place the home near her mouth

Difference of Field Experiments with Field Studies

1.       Field Experiments - these are experiments conducted in a real-life setting

2.       Field Studies - these are non-experimental designs used in real-life settings and include naturalistic observation, unobtrusive measures, participant-observer studies, and surveys

●       Paradigm Shift - a change in attitudes, values, beliefs, methods, and procedures accepted during a specific time period. This represents the increased use of qualitative research

●       Qualitative research is invaluable in studying contextual phenomena behavior that can only be understood within its context

○       We might examine the meaning of religious faith for patients facing impending surgery

Empirical phenomenology might rely on an experimenter’s private experiences or other experiential data:

1.      The researcher’s self-reflection on relevant experiences

2.      Participants’ oral or written descriptions of their experiences

3.      Accounts from literature, poetry, visual art, television, theatre, and phenomenological (and other) research

●       We can efficiently collect large amounts of data.

●       Anonymous surveys can increase the accuracy of answers to sensitive questions.

●       Surveys can allow us to draw inferences about the causes of behavior and can complement laboratory and field experiments.

●       The survey approach does not allow us to test hypotheses about causal relationships because we do not manipulate independent variables and control extraneous variables; this is the important limitation of the survey approach

1.      Identify specific research objectives

2.      Decide on the degree of imposition of units (Degree of response restriction: free or restricted?)

3.      Decide how you will analyze the survey data

1.      Keep items simple and unambiguous and avoid double negatives

2.      Avoid double-barreled (compound) questions that require responses about two or more unrelated ideas

3.      Use exhaustive response choices

●       In selecting the appropriate measurement scale to utilize in a study: the best type of scale depends on the variable you are studying and the level of precision you desire

○       Psychological variables like traits, attitudes, and preferences, represent a continuous dimension, and several levels of measurement “fit” equally well

○       When working with variables like sociability, psychologists often select the highest scale since it provides more information and allows analysis using more powerful statistics

➔     Subjects decide to refuse to answer surveys during the start or first few questions.

➔     Engage subjects from the start by asking interesting questions they will not mind answering.

➔     The first survey question should be:

1.      Relevant to the survey’s central topic

2.      Easy to answer

3.      Interesting

4.      Answerable by most respondents

5.      Closed format

➔     Whenever possible, use commonly used response options.

➔     Avoid value-laden questions that might make a response seem embarrassing.

◆      Version 1: Do you believe doctor should be allowed to kill unborn babies during the first trimester of pregnancy

◆      Version 2: do you believe doctors should be allowed to terminate a pregnancy during the first trimester?

Two Advantages of Probability Sampling (over nonprobability sampling):

1.      A probability sample is more likely to represent the population (external validity) than a nonprobability sample.

2.      We know the exact odds of members of the population being included in our sample. This tells us whom the sample represents.

1.      Simple Random Sampling

2.      Systematic Random Sampling

3.      Stratified Random Sampling

4.      Cluster Sampling

1.      Quota Sampling

2.      Convenience Sampling

3.      Purposive Sampling

4.      Snowball Sampling