6/27/2012

• Science is based on principles of:

– Objectivity: Scientists are expected to prevent their
personal biases from influencing the interpretation of
their results (Weber’s concept verstehen). Data is to be
interpreted on the basis of merit
– Verifiability: A study can be repeated by others and
allows critical analysis, retesting, and revision by
colleagues
• Methodology
– Quantitative Research: Collects and reports data
primarily in numerical form.
Professor David M. Long – Qualitative Research: Relies on what is seen in field
[email protected] and naturalistic settings; often focuses on small groups,
subcultures, & communities.

Research
Question • Variable: Measurable trait or characteristic
subject to change under different
Developing
conditions
Research
Conclusions Design – Quantitative Variables: Variation that can
Theory be measured and given numerical value
(e.g., individuals, groups, objects)
– Qualitative Variables: Variation in kind
versus in numerical degree (e.g., sex, marital
Data Data status)
Analysis Collection

• Independent Variable: Variable hypothesized to • Control Variables: Factors held constant

cause or influence another. to test the impact of the independent
– The factors that are manipulated by the experimenter. variable.
– The topic of research has nothing to do with their
occurrence. • Intervening Variables: Influences the
• Dependent Variable: Variable in which a change relationship between an independent
is expected to occur. variable and dependent variable.
– The anticipated change depends on the influence of – Hawthorne Effect: Unintended influence of
the independent variable. observers or experiments on subjects
– There may be several dependent variables.
– Dependent variables are often considered the scores
or outcomes of a study.

1
 6/27/2012

• Research Design: Detailed plan or • Experimental & Control Groups: Randomly

method for obtaining data scientifically assign research participants to either the
experimental group or the control group.
– Experimental
– The experimental group is the group of
– Correlational
subjects that will be exposed to the
– Survey independent variable.
– Observational – The control group will not be exposed to the
– Archival or Meta-analysis independent variable.

• It is not always possible to perform experimental designs

to test hypotheses, e.g., love, mob behavior, divorce.
• Since these topics are still very interesting, researchers
use a less powerful type of design, a correlation.
• With a correlational design you cannot draw cause-and-
effect conclusions (if you used an experimental method
you could have drawn cause-and-effect conclusions).
• A correlation is a measure of the relationship between two
variables taken from the each member of a group.
– Positive Correlation: Both the independent and dependent
variable(s) change in the same direction
– Negative Correlation: Independent and dependent variable(s)
change in opposite directions (when one goes up the other
goes down)

• Correlation: Exists when a change in one variable • Survey: Study that provides sociologists with information
coincides with a change in another variable. about how people act or think
– Interview: Researcher obtains information through face-to-face or
– Correlation does not necessarily indicate causation. telephone questioning
– Criteria for causality: – Questionnaire: Researcher uses printed or written form to obtain
• Two variables must be correlated information from respondent
• All possible contaminating factors must be taken into account • Return rate is often low (around 30%).
• A change in the independent variable must occur before a • Your sample is no longer random because often only the same type of
change in the dependent variable can occur. people (e.g., same economic status) return your questionnaire.
• Answers may be suspect. Person can not be truthful.
– Spurious Correlation: A mathematical relationship in
• It is extremely difficult or impossible to follow up on responses or delve
which two occurrences have no causal connection, yet it deeper.
may be inferred that they do, due to a certain third, unseen
– Surveys are often used when a research wants to study a large
factor (referred to as a "confounding factor"). number of people.
• The spurious relationship gives an impression of a significant – Survey research is often only descriptive in nature.
link between two groups that is invalid when objectively
examined – Surveys are used to develop theories and hypotheses.

2
 6/27/2012

• Reliability: Accuracy, precision, or consistency of

measurement.
– Degree to which measures are free from error and therefore yield
consistent results
• Would two people understand a question in the same way?
• Would the same person give the same answers under similar
circumstances?
– Reliable measures mean the same data would have been collected
under similar circumstances
• Validity: Degree to which a study accurately reflects or
assesses the specific concept that the researcher is
attempting to measure
– While reliability is concerned with the accuracy of the actual
measuring instrument or procedure, validity is concerned with the
study's success at measuring what the researchers set out to
measure
• Do the findings reflect the opinions, attitudes, and behaviors of the
target population?

• Case histories are often very interesting research • Systematic study of behavior in which the researcher
methods in which intensive information is is immersed in the social setting of the subjects.
– Pros: You can see the subjects in natural settings, and this
collected about one or a few individuals. The
can lead to the development of theories or hypotheses.
information collected can include, test scores, – Cons: You cannot control any variables, thus you cannot
medical and demographic data, subjective draw cause-and-effect conclusions. The results you obtain
accounts, or anything of interest. are very descriptive in nature.
– Pros: Can provide the basis for hypotheses • Types of field studies:
development and subsequent controlled research. – Participant Observation: Sociologist joins group to get
– Cons: Relies on the accuracy and objectivity of the accurate sense of how it operates
person reporting. It is also difficult to generalize from a – Ethnography: Efforts to describe an entire social setting
case study to the population at large. And case studies through extended systematic observation
are very difficult to replicate.

• Secondary Analysis: Research techniques

that make use of previously collected and
publicly accessible information, such as letters.
• Content Analysis: Systematic coding and
objective recording of data, guided by some
rationale
• Meta-Analysis: Combination of different
researchers’ data sets.

3
 6/27/2012

The
Scientific • Before any type of research can begin, you need
Method an idea of what is the problem and how it can be
defined.
– This is often facilitated by theories, which are broad-
based scientific frameworks that help to tentatively
explain the causes of behavior.
– Theories are made up of sets of statements that help
to explain a particular phenomenon.
– Theories help to:
• Organize data
• Provide direction

• Literature Review:
Theory – The most difficult part of any type of research.
– You want to know what other people have done.
• Hypotheses: Tentative predictions or educated guesses
about the outcome of a study.
– Hypotheses are based on the literature review, and are more
Generalization Hypothesis specific than theories.
– Can be described as the 'true' starting point of your research
study.
• Operational Definitions: Once your hypotheses have
been stated, you now describe exactly what variables
Data you will be observing.

• Population: After identifying your variables, you have to

determine all of the members of the group that you want • Random Sample: Each member of your
to study. population has an equal chance of being
– This is the group of people that will ultimately be affected by your selected (e.g., pick names out of a phone book
findings.
– The population you choose is usually too large to test everyone
or out of a hat).
that comprises the group. • Convenience Sampling: Members of the
– Because it is impossible to measure everyone in your population, population are chosen based on their relative
you must choose and measure a sample of the population.
ease of access.
• Sample: A subgroup of the population.
– How large samples should be?
• Large enough to be representative and small enough to be
manageable.
• To ensure a representative sample, your sample must be selected
randomly.

4
 6/27/2012

• Systematic Sampling: In a population of n units • Stratified Sampling: Method used to divide a

and a desired sample of s units, the auditor population into homogeneous subgroups (strata),
selects every r th unit (n/s) systematically and each stratum is then sampled individually.
beginning at a random point among the first r
• Cluster Sampling: The population is subdivided
units in the population.
into groups called clusters so that there is small
– Using this procedure, each element in the population
has a known and equal probability of selection.
variability within clusters and large variability
– The researcher must ensure that the chosen sampling
between clusters.
interval does not hide a pattern. – The first unit of each group is selected with the use of
• Any pattern would threaten randomness. a random number table.
• A random starting point must also be selected.
– This allows selection of more than one item at a time.

• Null Hypothesis: Describes in a formal way some • In statistical hypothesis testing, the p-value is the
aspect of the statistical behavior of a set of data and probability of obtaining a result at least as extreme as the
this description is treated as valid unless the actual one that was actually observed, given that the null
behavior of the data contradicts this assumption. hypothesis is true.
• In modern investigation the generally accepted value for
• Probability: A way of expressing knowledge or belief
significance must have a probability of occurrence by
that an event will occur or has occurred.
chance factors equal to or less than five times in 100 (p <
• Significance: A difference of such magnitude 0.05).
between two statistics, computed from separate • Generally, one rejects the null hypothesis if the p-value is
samples, that the probability of the value obtained will smaller than or equal to the significance level.
not occur by chance alone with significant frequency • If the level is 0.05, then the results are only 5% likely to be
and hence can be attributed to something other than as extraordinary as just seen, given that the null
chance. hypothesis is true.

 Percentage: Shows portion of 100 • Bias: An unintentional error in a particular

 Mean: Average; sum of a series of values direction that may produce misleading or
divided by the number of values erroneous conclusions.
• (1+2+2+3+4+7+9) / 7 = 4 – Expectations of subjects. Often they believe they
 Mode: Single most common value in a know what the experimenter wants, so to
series of values accommodate them, they answer questions in a
particular direction instead of being honest.
• 1, 2, 2, 3, 4, 7, 9 = 2
– Expectations of the experimenters. While
 Median: Midpoint that divides a series of experimenters are suppose to open-minded and
values into two groups with equal neutral, sometimes they can believe in something so
numbers of values much that they may see something that really is not
• 1, 2, 2, 3, 4, 7, 9 = 3 there, or may interpret some neutral behavior in a
way that confirms their theory or hypothesis.

5
 6/27/2012

• Researchers can prevent bias by

incorporating a single-blind or double-blind • Code of Ethics (American Sociological Association, 1997)
approach. 1. Maintain objectivity and integrity in research
– In a single-blind study the subject does not 2. Respect subjects’ right to privacy and dignity
know if they are in experimental or control 3. Protect subjects from personal harm
group. However, the researcher does know 4. Preserve confidentiality
which subjects are being exposed to the
5. Seek informed consent
independent variable.
6. Acknowledge collaboration and assistance
– In a double-blind study neither the subjects
nor the experimenter knows who is in 7. Disclose sources of financial support
experimental group.