Lecture Three Flashcards
Quantitative Research & Sampling
Unit of Analysis:
In research, the group of people, things, or entities that is being investigated or studied. For example, in organizational contexts, data can be collected from students, who in turn are part of schools, which in turn are part of districts, which may have multiple sites in several counties.
Operational Definition:
a description of something in terms of the operations (procedures, actions, or processes) by which it could be observed and measured. For example, the operational definition of anxiety could be in terms of a test score, withdrawal from a situation, or activation of the sympathetic nervous system. The process of creating an operational definition is known as operationalization.
Variables
A variable is an inconsistent figure that changes over time, between groups, etc.
-Gender
-Personality traits
-Income in a population
Variables may be discreet, having a finite range, or continuous, with infinite possible values between each observation. Discrete variables may have two (dichotomous) or more (polytomous) possible values.
independent variable
the variable that is manipulated by the researcher. This may be one or more treatment conditions manipulated prior to observing changes in the dependent variable.
mediator variable
part of the process or mechanism by which an independent variable produces outcome on the dependent variable.
moderator variable
influences the direction or strength of the relationship between the IV and the DV, but does not have a causal connection from the independent variable
Scales of Measurement
The four scales of measurement describe which property or set of properties a phenomena adheres to, and thus determines what statistics we can use to describe and summarize the construct.
Nominal, ordinal, interval, ratio.
See table from slides
Nominal
describes arbitrary labels used to identify individual categorizations (identity). Examples would include social security numbers, zip codes, or arbitrary numerical codes in research.
Examples:
- The number on an athlete’s uniform
- Your social security number
- Your Visa card number
- The city where you live
Ordinal:
used to order a hierarchical series, such as rank order (first, second, third…), or percentiles. Each value is greater or less than another (magnitude), though no standard interval exists between values. For example, the difference between first and second place may not be the same as the difference between second and third.
- Order of finish in a race or -a contest
- Letter grades: A, B, C, D, or F
- Level of agreement (e.g., Likert scale): Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree
Interval
contains an ordered series with equal intervals between values, but no absolute-zero point. Consider temperature in degrees Fahrenheit, a measure of heat. The difference between 10° and 15° is the same as between 90° and 95°, but a temperature of 0° does not indicate a complete absence of heat. IQ and SAT scores are interval scales.
Examples:
- Scores on the College Board’s Scholastic Aptitude Test, which measures a student’s scores on reading, writing, and math on a scale of 200 to 800
- Intelligence Quotient scores
- Dates on a calendar
Ratio:
an ordered series with equal intervals between values and an absolute-zero or “starting” point. Distance from a starting point or temperature in degrees Kelvin are both ratio scales, as a score of zero indicates an absolute zero, or the complete absence of the construct being measured.
Examples:
-Height
-Income
-Distance travelled
-Time elapsed or time remaining
Descriptive statistics
procedures used to summarize, organize, and simplify data, and may be used with either a sample or a population.
-Central tendency: mean, median, mode
-Variability: variance, standard deviation
Inferential statistics
techniques that allow us to study a sample and then make generalizations about the population from which the sample is drawn. compares variables to each other and generalizes to the population
- T-test
- ANOVA
- Correlation
Population
refers to the entire set of individuals of interest in the study.
sample
the set of individuals selected to represent the population.
Sampling in the modern world
-In-person
-Phones and call or polling centers
-Web research: email, websites, consumer surveys, Amazon Mechanical Turk and apps
probability sample
each member of the population has a CHANCE of being selected.
simple random sample
each member of the population has an EQUAL chance of being selected.
external validity
sometimes referred to as validity generalization.
Choosing a probability sample allows for generalizations to the population, thereby increasing external validity
Probability samples:
- Simple random
- Cluster: groups of participants, usually location-based
- Stratified: proportional, random sample of homogeneous groups
- Systematic (every nth)
- Oversampling
Nonprobability samples
better for direct exploration of topics in greater depth, qualitative studies
- Convenience
- Purposeful: individuals judged to be good sources of information
- Snowball: participants rec. other participants
- Self-selection
Sample Size
- Large n increases precision; law of large numbers
- Can start with pilot study to check
- From a few hundred for a SBS article to 1,500 or more for national studies
- Minimum 100; 300+ good; 10% rule of thumb
- Depends on the number of variables, group heterogeneity and variance, and confidence
- Sample size calculators: Qualtrics
Qualitative Samples
- Smaller, average approx. 20-50
- Studies are more reliant on researcher’s judgment than statistical analysis
- Not trying to generalize to population in many cases
- Saturation: when additional responses no longer add variability to the results being collected (i.e., no new themes, variables, or relationships emerge)
Standardized vs Norm Referenced Scores
Standardized: everything is the same across the board
Norm Referenced: scores on a test are adjusted compared to the group
IQ Score
Mean of 100, standard deviation of 15
Norm referenced
If someone gets an IQ score of 100 regardless of age it is spot on average.
Law of Large Numbers
Outliers are less likely to skew data with a higher sample size and can help decrease systemic bias (but will not save you from it)
*check this definition plz