Lecture 5 Flashcards
Definition of THEORY
interrelated set of constructs (variables)
put into propositions or hypotheses (predictive statements)
specifies variable relationships
tested
to give a scientific explanation or prediction (what the researcher expects to find)
can be series of “if then” statements
Quantitative researcher uses a theory to… (3)
create a hypothesis
conduct a study
present his/her findings
Levels of Theory (3)
Micro (interactions between people)
Meso (organizational, social movements, or community)
Macro (social institutions, cultural systems, societies)
Deductive Approach
“top down” process
- researcher tests or verifies a theory
- researcher tests hypotheses or research questions from the theory
- researcher defines and operationalizes variables derived from the theory
- researcher measures or observes variables using an instrument to obtain scores
Variables
characteristic/attribute of an individual or organization
measurable/observable
varies
common examples: age, gender, SES, attitudes, behaviours
sometimes referred to as constructs
6 different kinds of variables
Independent Dependent Intervening/mediating Moderating Control Confounding
Independent Variable
probably cause/influence/affect outcome
may be called: treatment, manipulated, antecedent, or predictor variables
Dependent Variables
depend on independent variable
outcome/result of the influence
Intervening/mediating variables
between independent variables and dependent variables
mediates the effects of the independent variable
Example of Intervening and mediating variables
Study prep (Independent Variable) … Organization of Ideas (Intervening variable) … Test performance (Dependent variable)
Moderating Variable
affect direction +/or strength of relationship
new - constructed by researcher
1 V X another V to determine joint impact
example: age x quality of life attitudes impacting self esteem
Control Variables
type of independent variable (eg. age, gender)
controlled via statistics to try to find out actual impact of independent variable on a dependent variable
Confounding (spurious) variables
not measured in the study (difficult to assess)
commented on afterwards
Surveys
provides a numeric description of trends, attitudes, opinions
studies a sample (then generalizes to population)
Experiments
tests impact of a treatment or intervention on outcome
controls for other factors that may influence that outcome
sample -> population
Numeric description
trends, attitudes, opinions
study a sample of a population
Advantage of Surveys
quick turnaround and large sample possibilities
Forms of Data Collection
mail telephone internet personal interviews group administered
Single stage of a sampling design
access to names with a population/sample directly
Multi-stage (clustering)
organizations -> names within organizations -> sample
Selection Process (3)
Random (equal probability of being selected)
Non-probability (convenience sample/available)
Stratification (eg. gender, age equally represented)
Population vs Sample
Example: UNBC
Population: all UNBC students
Sample: # of students included in survey
3 Survey Instruments Choices
- Specifically design instrument for study purposes
- Intact, existing instrument
- Modified existing instrument
Survey questions
relating variables to research questions to item on survey
closed
two scales used: continuous & categorical
Mail out process of a survey
advance notice, survey,
follow up: 4-8 days later,
4th notice to non-respondents: 3 weeks later
Response Rate
how many respondents did the survey have?
Response Bias
what is effect of non-responses?
would these non-responders have affected results?
Descriptive Stats
frequency, mean, SD, range
Inferential Stats
various tests for significance (t-test, analysis of variance, Chi-Square)
Interpretation
drawing conclusions from results
statistically significant or not
Key Ideas of Surveys
cross sectional longitudinal population - sample questions instrument used for data collection procedures for distribution/follow up statistical analysis
Experiments
tests impact of a treatment on an outcome
controls for other factors that might influence (eg. random selection of individuals to groups)
one group gets treatment, other does not
measures difference
What do experiments define?
Independent, dependent, intervening variables
4 Designs of Experiments
- pre-experimental
- quasi-experimental
- true experiments
- single subject
Pre-experimental
single group, intervention studied, no control group
Quasi experimental
participants not randomly assigned to groups
True experiments
participants randomly assigned
Single-subject
observing single individual/small number over time
