Comprehensive Guide to Variables, Relationships, and Qualitative Methods in Research

CONTINUOUS VARIABLES

(infinite number of values or attributes)

ATTRIBUTES along a continuum for a variable

Can be made discrete

DISCRETE VARIABLES

(fixed set of values or attributes)

Definite number of attributes attached to variable

Can not be made continuous

NOMINAL

(categories) labels or names only

ORDINAL

(rank order) one is greater than another

INTERVAL

(distance) arbitrary 0

RATIO

(proportion) absolute 0

leveles of measurement/variables

higher levels include the lower levels

CAUSAL RELATIONSHIP

relationship between two events

three requirements

TEMPORAL ORDER/CAUSALITY

(cause before effect)

ASSOCIATION

(some relationship between variables, but not necessarily CORRELATION (definite and strong cause and effect)

can have association without

causality

FREQUENCY DISTRIBUTION

(main characteristics of participants in the sample)

Display in a table, graph, or chart

CENTRAL TENDENCY

(center point of distribution)

Normal bell-shaped curve

VARIABILITY

(dispersion/spread of distribution)

Same mean, meadian, and mode may have different distributions

Results from one variable

mean, median, mode, range, percentile, and standard deviation

MEAN (central tendency)

(average)

Impacted by extremes

MEDIAN (central tendency)

(middle number/mid-point)

50% above and 50% below

MODE (central tendency)

(most common/frequent number)

RANGE (variability)

(largest and smallest scores)

PERCENTILE (variability)

(score at a specific point)

STANDARD DEVIATION (variability)

(measure of dispersion; distance between all scores and the mean)

measure two variables

bivariate relationship, scattergram/scatterplot, and bivariate table

BIVARIATE RELATIONSHIP

(relationship between two variables)

covariation

independence

COVARIATION

(an association exists; variables vary together)

INDEPENDENCE

(no relationship exists)

scattergram/scatterplot

(graph)

Form (whether relationship exists)

Direction (positive/negative relationship)

Precision (spread in the points on the graph)

Bivariate table

visually difficult

measure three or more variables

multivariate relationship, partials (trivariate tables), multiple regression

MULTIVARIATE RELATIONSHIP

(relationship among three or more variables)

make sure spuriousness does not exist

"Control for" one or more variables (no effect)

if the bivariate relationship is weakened by excluding the "control for variable" then the "control for variable" is important. this means that te bivariate relationship is spurious

Partials (Trivariate tables)

Bivariate tables for independent and dependent variables

STATISTICALLY SIGNIFICANT

(results are not due to chance)

Sample may be different than population

Likelihood (not absolute certainly)

Two variables my be statistically significant without a relationship

LEVEL OF SIGNIFICANCE

LEVEL OF SIGNIFICANCE

(probability)

.05 means 5 in 100times will be due to chance; 95 in 100 times will not be due to chance (95% confident)

TYPE I ERROR

Researcher says relationship exists between the variables, when there is no relationship

TYPE II ERROR

Researcher says relationship does not exist between the variables, when there is a relationship

Types of Errors

Type I and type II

.05 level of significance balances types I and II errors

Cautious researcher uses .01 level of significance (1% due to chance; 99% confident

Riskier researcher uses .10 level of significance (10% due to chance; 90% confident)

Conceptualization

Data collection and anaylsis

New concepts formed

Concepts refined

CASING

(create/justify a case so data connects with theory)

OPEN CODING

(initial categorizations based on commonalities)

AXIAL CODING

(analyze the initial codes to create themes/sub themes)

SELECTIVE CODING

(focus on selected themes that support conceptual framework)

Analytic strategies

narrative history, successive approximation. illustrative method, and ideal types

Max Weber

NARRATIVE HISTORY

(tell a story in chronological order)

Authenticity (views of participants)

SUCCESSIVE APPROXIMATION

(collect additional information)

Back-and-forth between data and theory

ILLUSTRATIVE METHOD

(apply theory to data)

IDEAL TYPES

(perfect/exaggerated models and standards to compare reality)

CONTRAST CONTEXTS (uniqueness)

ANALOGIES (similarities)