CONDUCTING RESEARCH IN SOCIAL PSYCHOLOGY RESEARCH PROCESS, DESIGNS AND METHODS (CONTINUED)

<< CONDUCTING RESEARCH IN SOCIAL PSYCHOLOGY:Evaluate Ethics

CONDUCTING RESEARCH IN SOCIAL PSYCHOLOGY OBSERVATIONAL METHOD >>

Social Psychology (PSY403)

Lesson 05

CONDUCTING RESEARCH IN SOCIAL PSYCHOLOGY RESEARCH PROCESS, DESIGNS

AND METHODS (CONTINUED)

Aims:

To make students understand how to conduct social psychology research using diverse scientific

methodologies.

Objectives:

Describe empirical research and scientific method

Discuss basic steps in conducting social psychology research

Discuss important research designs

Describe observational research method

RESEARCH PROCESS

Step 6: Collect Data

In the previous lecture relationship between population and sample was described. Regarding different

sampling strategies, probability sampling, where each and every member of the population has an equal

chance of being included, was discussed. Hence, this lecture will start from Non-probability sampling

techniques, which can not guarantee the representativeness of the sample of its base population.

Non-probability samples: May not be representative; validity of inferences about population

characteristics is neither assumed nor testable

1. Purposive sampling: the investigator selects the elements to be included in the sample on the basis

of their special characteristics.

2. Accidental sampling: involves selecting respondents primarily on the basis of their availability and

willingness to respond, like newspaper interviews of people on the streets, their comments are

interesting but not necessarily representative of the wider community:

3. Convenience sampling: group composition is heavily biased and results shall never be generalized.

Commonly used in pilot studies.

4. Call-in surveys: usually used by TV and radio shows to poll the views of their audience. People

who are willing to call, pay the calls make the sample. Crossen (1994) compares the results

obtained by a random sample survey, and a call-in survey. The question was, "should UN

headquarter stay in the USA or not". In the Call-in survey, a solid majority, 67%, of 1,86,000

callers replied that it should stay out of USA, while a clear majority, 72% of survey subjects

wanted in the USA.). Despite a massive sample size, the call-in survey can not be considered

representative.

Types of Data

Qualitative data exists in a nonnumeric form

Quantitative data is numerical.

Techniques of data collection

· Self reports: Useful to measure important subjective states, e.g., attitudes, emotions, and

perceptions. However, they rely on people's own descriptions of internal states. Because of this

drawback some people prefer direct observation

· Direct observations

· Archival information (accumulated records): court records, newspaper (my studies), census, govt.

documents)

These techniques are not mutually exclusive

Social Psychology (PSY403)

7. Analyze the data and Reevaluate the theory

Once the data have been collected, the first part of the 7th step is to conduct data analysis, which usually

requires extensive knowledge of statistical procedures and computer software packages. The two basic

kinds of statistics are descriptive and inferential.

Types of statistics

· Descriptive Statistics merely summarize and describe the behavior or characteristics of a particular

sample of participants in a study.

Inferential statistics move beyond mere description to make inferences about the larger population

from which the sample was drawn.

After data analysis determines whether the hypotheses successfully predicted the outcome of the study,

researchers next reevaluate the theory. Were the research hypotheses supported by the data, which thereby

supports the validity of the theory? If the data do not support the study's hypotheses, or if only some of the

hypotheses were supported, the theory probably needs revising.

Example: Hoveland and Sears (1940) found significant correlation between economic frustration and

lynching of black by white people in Old South.

Analysis of data also determines the validity of theory.

Step 8: Report the Results

"If research was not "written up," did it really occur?

Three ways to disseminate knowledge:

· Publishing articles in scientific journals

· Making presentations at professional meetings

· Personally informing other researchers

Parts of a research article

· Title

· Authors and their affiliations

· Abstract

· Introduction

· Method

· Results

· Discussion

· Conclusions

· References

· Tables

· Figures

· Appendices (if any)

· Footnotes (including Author Note)

Research Designs

It is almost always helpful to make a firm distinction between designs based on looking for differences

between groups and those looking for relationships/ associations between variables (Cohen & Holiday,

1982). This distinction makes the subsequent choice of appropriate analyses and statistics much more

straightforward.

Important Research Designs

There are 3 important research designs.

Social Psychology (PSY403)

1. Descriptive investigations: Non-experimental designs geared towards describing phenomena. e.g., a

quantitative study to determine the normal range of body temperature of a population; a qualitative study of

the environment in an intensive care unit.

2. Designs based on differences between groups

· Independent/ unrelated groups: ideally created by random independent sampling, or by post-hoc

division.

Dependent/ related groups: groups linked by some factor which means that the values of one

group are likely to influence the others (correlated); a clear example is the use of repeated-

measures design to assess change over time. Repeated-measure are very popular designs as they

require less subjects as they act as their own control. Moreover, within subject variability stays the

same under all repeated conditions. However, they have problems of order effect, which can be

overcome by counterbalancing, a strategy when half of the group is tested before (in condition 1),

and the other half is tested first in the 2nd condition, and the order is reversed for both groups. A

similar design to repeated/ dependent group design is matched samples design, where subjects are

so closely matched across conditions that they are related.

3. Designs based on association between variables: Also used in experimental situations when the level

of a variable is manipulated and associated change in the other variable is investigated, e.g., effect of

punishment and reward on performance.

Extensions of earlier 3 designs:

Longitudinal designs: Really an extension of a simple repeated measure design; subjects or events

are studied in order to investigate change over prolonged periods of time, as in some case studies. It

may also be called a kind of extended dependent group design.

Mixed designs are combinations of independent and dependent group designs, hence having the

characteristics of both.

Most Common Research Methods

The most commonly used research methods used in social psychology research are:

Observational research

Correlational research: Surveys

Experimental Methods

Observational Research

Description Is the Goal of Observational Research

To understand behavior so that it can be predicted, controlled, or explained, a scientist must first describe it

accurately. Observational research is a scientific method involving systematic qualitative and/or

quantitative descriptions of behavior. In collecting this data, the scientific observer would not try to

manipulate (that is, change) the behavior under study but would simply record it. Description is the primary

goal here. The common types of observational methods employed by social psychologists are naturalistic

observation, participant observation, and archival research.

Social Psychology (PSY403)

Characteristics of observational research

Simply record the behaviour, don't attempt to change it

High in validity: as the observed behaviour are real and actual sample of target behaviour rather

than a sign of behavior .e.g., observation of pro-social behaviour at the time of real accident on the

road.

High in resource consumption: many apparatus or instruments may be required, time-consuming,

difficult to quantify

Essentials of Observation

1. Functional Analysis (Skinner, 1972) is an essential of observation consisting on:

Antecedent

Behavior

Consequence

2. Categories of behaviour

Categories of target behaviour need to be very clearly described and then operationally defined for better

understanding and measurement. For example, Interpersonal interaction could have following further

categories: initiates conversation, maintains conversation, maintains eye contact, among others, etc.

Categories of Aggressive behaviour:

Physical or verbal?

Physical: Pushing, hitting, kicking, punching, pinching, biting, throwing things, slapping, etc.

Verbal: Staring, frowning, cursing, speaking in a threatening way, name calling, etc

Readings

2. Franzoi, S.L. (2006). Social Psychology. New York: McGraw Hill. Chapter 2.

Table of Contents: