Classical scientific theory begins in general theory and uses deduction to frame its research following the above cycle. Grounded theory backs up specific observations as the beginning point using induction to create general theory. In operational theory, the researcher considers the meaning of a concept to be a set of operations. Therefore, the process of operationalizing is the core of all research and its subsequent results in interpreting observations.

RESEARCH METHODS: A systematic strategy for conducting research

A. Steps in Social Research

1. Choose the research topic
2. Review existing information on the problem
3. Formulate variables and hypothesis
4. Choose research design experiments

• survey research
• field research
• content analysis
• existing data, i.e. historical research, comparative research

1. 5. Operationalize the variables in the hypothesis.
2. 6. Develop research questions and gather data.
3. 7. Code and analyze data
4. 8. Interpret the results
5. 9. Report or publish results

B. Considerations In Choosing Research Topic

1. Sociological paradigm: a cohesive group of basic assumptions that frames thinking and research.

• structural functional: assumes society is a system of parts that work together to promote stability.
• conflict: assumes society is characterized by inequality & conflict that generate change.
• symbolic interactionalism: assumes society is the product of the countless everyday interactions of individuals.

1. Researcher Values: There are two schools of thought. Max Weber proposed what he called "value-free" research. Although there is a concession to the premise that researchers choose areas of research that involve themselves in someway, Weber said that their subjective interests should stop there. Researchers should ideally pursue all subsequent knowledge with an objective "disinterested" approach.
   Alvin Gouldner later countered that this did not describe most research. He contended the purpose of social research is to evaluate and explain society with the intent of changing and improving. Therefore, although researchers need to objectively apply their skills and analysis, they are by no means required to be "disinterested" and uninvolved with the political ramifications of the research, or refrain from giving their informed support.
   

C. Considerations In Choosing A Research Design

1. Purpose of research: explain, predict, describe
2. Degree of reactivity inherent in the methodology used.
3. Limitations and scope of the study
4. How time is treated?
   • cross-sectional: data collected represents a single point in time.
   • longitudinal: data collected over different points in time i.e. trend studies (large populations over course of time),
   • cohort studies (specific populations over course of time),
   • panel studies (specific samples over course of time)

5. Type and source of sample

6. Validity: Does the research instrument measure precisely what it intends to?

7. Reliability of measurement: Can the stated concepts be measured consistently with replicable results?

DEFINITIONS

scientific knowledge: Knowledge based on empirical evidence (information verified or obtained through one's senses).

theory: A set of interrelated propositions explaining the topic being studied.

hypothesis: A testable statement indicating a relationship between variables.

concept: An abstract construct that connotes a part of the world, (i.e. family, society).

variables: A concept whose mutually exclusive values change from case to case.

measurement: The process of determining the value of a variable in a specific case.

operationalizing: Specifying exactly what one is to measure in assigning a value to a variable, (e.g. class may be measured by income, occupational prestige, education).

population: The members of a group who are the focus of the research.

sample: A part of a population researchers select to represent the whole.

qualitative research: Observations that are not easily reduced to numeric forms, emphasis on inductive thought and subjective analysis.

quantitative research: Observations that are analyzed statistically, emphasis on deductive thought.

SAMPLING METHODS

A. Definitions

• units of analysis: What is being studied, individuals, groups, gangs, families, organizations, social artifacts, books, films, etc.
• ecological fallacy: Making generalizations about individuals when the units of analysis are based on larger groups.
• sample frame: A list of populations being studied, i.e. a roster.
• parameter: Description of the variable in the population.
  

B. Types Of Sampling

1. random: Each element has an equal chance of being selected independent of any other event in the process.

• simple random: Each unit is assigned a number, use random tables. Units are then chosen.
• systematic: Begins at random, but have a sampling interval that is a consistent distance or interval between. The sampling interval equals the population size divided by the sample size.
• stratified: The population is sorted into elements of discrete groups on basis of relative proportion in the population. Then sample size is randomly selected from each group that reflects the same proportion. This ensures proper representation.
• cluster: When it is not feasible to complete an exhaustive list, larger units (i.e. sample blocks) are first randomly chosen then a second random selection using smaller units (households) is made. List the households, sample the household, list the persons, sample the persons. This is subject to two sample errors, because of the repetitive process.

*NOTE: sampling error reduced by 2 factors: sample size & homogeneity

• probability proportionate to size: Used when clusters are greatly differing in size. Elements within each cluster are a fixed number (5 houses per block, etc.)

2. non-random

• quota: Create a matrix with proportions (e.g. affirmative action), then fill them based on those characteristics
• convenience: An easily obtained non-random sample.
• snowball: The subjects which meet the sample parameters are asked to give additional names. These people are then interviewed (sampled) and asked for names etc.
  

SURVEY (see survey example)

A. Types of Questions:

1. open ended which increases accuracy but is more difficult to code. Also, considerable time & effort is spent by the respondent.
2. closed ended which are discrete categories with no overlapping in the responses. Both types of survey questions ask about attitudes, beliefs, behavior, and characteristics of respondents.

B. Format Considerations

1. Make it as easy as possible. Reduce hindrances that may lower response rates. This includes difficult questions. At the start of the interview have simple descriptive questions. Then gradually bring in attitudinal questions, phasing the more controversial in a manner that does not bring emotion into the interview.
2. Present the survey form in as professional a manner as possible. This includes hiring interviewers which are trained and elicit respect through their appearance an professionalism. Also any accompanying paperwork or aides should match the content goal of highest quality.
3. Makes sure the directions are clear.
4. Include a cover letter introducing the interviewer and stating the purpose of the study. Be sure to include backing organizations which the interviewee will recognize.
5. Make responding as easy as possible. Keep the answers succinct. Make sure any multiple choice formats do not overlap. Use simple wording with direct questions. Avoid double negatives (e.g. Are you opposed to people who are not in favor of…
6. Be accurate in any statements that are made-particularly when quoting percentages.

After the data is gathered it is necessary to enter it into an database format. This book will use SPSS. An example of what the data would look like is presented below:

1. Example of an SPSS data file

Sex	Race	House Hrs Work/week	Employed	Hrs Worked per week	Hrs Study per week	Marital Status
1	1	5	0		18	1
1	0	1	1	8	31	0
1	0	2	1	10	13	0
0	1	6	1	30	20	1
1	2	3	0		12	0
1	0	7	1	30	3	0
1	0	10	0		12	0
1	3	8	1	12	10	0
1	1	10	1	20	9	0
0	2	3	1	26	4	0
0	0	3	1	25	21	0
1	3	3	0		99	1
0	3	14	1	27	8	1
1	0	2	0		40	0
0	2	2	0		20	0
1	2	6	1	30	15	0
1	0	1	0		5	0
0	0	15	1	42	20	0
0	3	2	1	13	6	1

The General Social Services (GSS)

GSS is part of the National Data Program for the Social Sciences at the National Opinion Research Center (NORC), University of Chicago. This survey attempts to poll American opinion through the use of standardized questionnaires conducted every year in February, March or April from 1972 to the present (except 1979 and 1981). In the original survey design, a modified probability with a quota element at the block level was utilized. Then in 1977, 78, 80, and 82 till present a full probability sample design was implemented. In 1983 and then again in 1993 a spit sample transition design was used to determine the effects of design changes with 1983 representing the 1970 to 1980 change and 1993 for 1980 to 1990. In 1983 and 1993, the variable Sample was added to indicate which split each represents. The survey takes about an hour and a half to complete and is done by professional interviewers. For further information, see the General Social Survey Cumulative Cookbook - Chicago: National Opinion Research Center; 1993.
next lesson

table of contents