Chapter 4 and 5: Research foundations and populations Flashcards
Indicator
means by which we assign individual cases to different values of the variables
Nominal Variable
Cannot be ordered or ranked
No mathematical relationship between the variables
numerical value represents differences in kinds, not in degree
Ordinal Variable
Organized into categories along a continuum
No precise distance between categories
Numbers represent relative positions
Interval/Ratio Variable
Can be ordered and categories are separated by a standard unit
Ratio variables can have an absolute zero
the distance between variables is the same
Which type of variable has the highest precision and which has the lowest?
Highest: interval/ratio
Lowest: Nominal
Theory-building research
Inductive
Seek to obtain real-world observations in order to create an explainable theory
Theory-testing research
Sets out to test the hypothesis established by a theory.
Inductive reasoning
using empirical evidence to draw a conclusion - to help form the definition of a concept
Operationalization
the process of moving from abstract concepts to concrete measurable variables
Correlation
A state in which two entities change in conjunction with the other.
Demonstrating that two concepts are correlated and have a relationship.
Causal mechanism
Explains which is the cause and effect relationship between variables.
A plausible explanation of why concepts are related
What is the difference between a causal mechanism and a correlation?
A causal relationship means there is one relationship to another.
Correlation simply states does not explain
What does a hypothesis identify? What is the opposite?
A hypothesis identifies a relationship between variables
A null hypothesis is the opposite of this. It means there is NO relationship between the variables.
Types of correlation?
Positive correlation: Both variables (IV AND DV) move in the same direction, meaning if one increases, so does the other
Negative correlation: The variables move in an opposing direction. Meaning if the iV increases, the DV would decrease.
Name examples for each type of variable
Nominal: Religious affiliation
Ordinal: Poor, fair, good, excellent
Interval: Temperature
Ratio: Income (absence of income)
Continuum
Ordering the values of a concept based on a dimension of low to high, less to more
What are the 5 criteria for having a strong causal argument?
Correlation
Temporal order: a shared pattern in the correlation. Time sequence of events.
Absence of confounding variables: eliminate any alternative explanation the relationship of variables can be accounted for with a third variable (spurious relationship)
Affects both IV and DV simultaneously
Plausible causal mechanism: Parosmonious (easy to explain) and direct
Consistency: replication of the study
Bivariate vs Multivariate
Bivariate: contains two variables
Multivariate: Contains multiple independent variables
Intervening variable
comes in between the IV and DV
What is the difference between a confounding variable and a reinforcing variable?
Confounding affects both, while reinforcing only affect the dependent variable
Categorical concepts vs continuous concepts
Categorical concepts: variations within the concept indicate differences of kind, and cases are categorized into groups according to descriptions
Continuous concepts: variations within the concept indicate differences in degree, and the concept’s characteristics are sequentially connected and categories are placed on a continuum.
Quantitative research
Uses numeric information
uses both categorical and continuous
Qualitative research
uses more textual information
whether a concept is categorical or continuous
Measurement error
the difference between the true value and the value obtained through measurement observed through a quantity
non-random vs random error
Random error: naturally occuring, no specific cause
Non-random error: faults in the measures, error in a specific direction. Creates bias in the data
Measure of validity and types
Whether the measurement accurately reflects the conceptual definition.
Internal and external
Internal validity: trustworthy assessment of causality, able to satisfy most if not all criteria of causality
External validity: the findings of the case may be used to make generalizations beyond the original study
Population
A group of people that researchers want to make a generalization about.
What are the 3 factors to consider when choosing a population?
unit of analysis, population parameter, sample statistic
Unit of analysis
focus group of the study. Defined. Geographic and temporal qualifications
Population parameter
Score of each member of the population is measured in numeric form
Resultant characteristic
Sample statistic
Score of a sample measured in numeric form
What is a sampling selection method and what are the two types
Manner by which cases in the population are selected for inclusion in the sample
Can be divided into two categories:
probability and non-probability
Probability sampling techniques are based on probability theory and allow researchers to use statistics to test the representativeness of their sample. Commonly used in quantitative research
Non-probability sampling: opt-in panel surveys. Participants make the decision to join the panel rather than being contacted through random selection and experimental research.
Sample size
number of cases involved in the sample
Probability samples and types
probability ranges from 0 to 1. 0 being no relationship and 1 being a 100% chance.
Random selection: Makes generalizations and pulls conclusions
Simple random sample: the process by which every case in the population is listed and the sample is selected randomly from this list
Systemic random sample: A selection interval is calculated based on the sample size needed. Following this, a random number is used as the selection starting point.
Stratified random sampling: Breaking the population into mutually exclusive subgroups, and then randomly sampling each group
Cluster Sampling: Process of dividing the population into a number of subgroups, known as clusters, and then randomly selecting clusters within which to randomly sample.
Non-probability samples and types
Accidental sample (convenience sample): researchers gather data from individuals they “accidentally” encounter or who are convenient
Ex: “person on the street”
Self-selection: respondents themselves select whether or not to be part of the sample
Ex: social media polls
Limited to those who opt in
Unrepresentative of a larger population
Purposive sampling: researcher selection of specific cases
Uses his or her judgement to select cases that will provide the most information
Used in comparative research
Most similar systems design: similar characteristics
Most different systems design
Snowball sampling: researcher begins by identifying a few cases and then getting referrals to others and continues to branch out
Potential for closed loops
Quota sampling: research identify a number of target subgroups, and then sets a quota number that must be met by each group
Design weights
To reconstruct a representative national sample, mathematical corrections to compensate for the fact that respondents probabilities of being selected were influenced by the research design,
sampling distribution
Sampling distribution: all the possible sample means for a given sample size. Created by totalling the number of combinations that present the specified sample mean
Confidence interval
Confidence interval: range of values within which the population parameter is likely to fall
Confidence level
level of confidence that the population parameter will fall within the confidence interval
What is the most typical level of confidence in social science?
95% confidence level
Margin of error
amount if sampling error, expressed as a percentage - they are willing to accept. Used in conjunction with the sample statistic plus or minus the margin of error
Sampling error
Sampling error: refers to the difference between the sample statistic and population parameter
A large sampling error indicates that the sample statistic deviates greatly from the population parameter
A small sampling error indicates the sample statistic is close to the population parameter
Proportionate stratified random sampling vs disproportionate
Proportionate stratified random sampling: generate a final sample that reflected the proportion of each set of students in the overall population
Disproportionate stratified random sampling: used if a particular group of interest is small. However, this may not reflect the general population anymore
