Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
244 Cards in this Set
- Front
- Back
empiricism
|
Knowledge comes from experience
|
|
Reasoning/Rationalism
|
idea that original knowledge comes from reasoning.
|
|
Deductive reasoning
|
process of drawing a conclusion that is necessarily true if the premises are true
|
|
Inductive reasoning
|
process of drawing a conclusion that is “probably” true
|
|
Science
|
an approach for generation of knowledge. Relies on empiricism (collection of data) and rationalism (use of reasoning and theory construction and testing).
|
|
confirmatory or deductive method
|
a form of scientific method.
1. State the hypothesis (based on theory or research literature), & deduce what must be observed if hypothesis is true. 2. Collect data to test the hypothesis. 3. Make a decision to tentatively accept or reject the hypothesis. is “top down” method for testing theories and hypotheses. |
|
Exploratory or Inductive Method
|
1. Observe the world in all of its particulars.
2. Search for patterns. 3. Make a descriptive conclusion or generalization. Exploratory method is commonly used by qualitative researchers. “bottom up” method for generating theories and hypotheses. |
|
Theory
|
"explanation."
Explains "How" and "Why" something operates as it does. |
|
Theoretical Paradigms
|
Frameworks for Understanding Phenomena
Systematic Approach to Analysis Tools for Problem Solving, Effective Engagement, and Reflective Practice Research Orientation to Study of Education |
|
Positivist
|
Theory and positivist research can be used to describe, understand and predict what happens in the world .
|
|
Social Constructionist
|
Theory and SC research helps us see how multiple realities are constructed, emphasizing perspective, communication, interpretation and sense-making in in the world
|
|
Post-Modern
|
Theory and P-M research helps discover hidden assumptions that guide behavior and phenomena and deconstruct these assumptions to address conditions of oppression and alienation in the world.
|
|
The Principle of Evidence
|
Empirical research provides evidence, NOT proof.
Research conclusions are tentative and probabilistic. Evidence increases when finding is replicated. Remember: It is normally difficult and problematic to draw firm conclusions from a single study. |
|
General Kinds of Research
|
basic research
applied research evaluation research action research orientational research |
|
Basic research
|
aimed at generating fundamental knowledge about natural processes.
|
|
Applied research
|
focused on practical questions; goal is to provide relatively immediate solutions.
|
|
Evaluation Research
|
determining the worth, merit, or quality of an evaluation object.
|
|
Formative evaluation
|
purpose of program improvement
|
|
Summative evaluation
|
purpose of making summary judgments to continue or discontinue program
|
|
Action Research
|
Focuses on solving practitioner’s local problems.
Conducted by practitioners. It’s a state of mind; teacher takes on research attitude, constantly testing new ideas. |
|
Orientational Research
|
Done for purpose of advancing an ideological position
Focused on inequality and discrimination Class stratification Gender stratification Ethnic and racial stratification Sexual orientation stratification International inequalities |
|
Paradigm
|
perspective based on set of assumptions, concepts, and values that are held and practiced by a community of researchers. 3 kinds:
Quantitative Qualitative Mixed |
|
Quantitative research
|
research that relies primarily on quantitative data (numerical).
Pure quantitative follows all of the paradigm characteristics of quantitative. |
|
Qualitative research
|
research that relies on qualitative data (non-numerical).
Pure qualitative research follows all of the paradigm characteristics of qualitative. |
|
Mixed research
|
mixing of quantitative and qualitative methods or other paradigm characteristics.
|
|
Variables
|
takes on different values or categories
|
|
Constants
|
single value or category of variable
|
|
Quantitative variables
|
vary in degree or amount (e.g., annual income)
|
|
Categorical variables
|
vary by type or kind (e.g., gender).
|
|
Independent variables ("IV")
|
presumed to cause a change in another variable.
|
|
Dependent variables ("DV")
|
presumed to be influenced by one or more independent variables.
|
|
extraneous variables
|
variables competing with IV in explaining outcome
|
|
Intervening variables
|
also called mediator or mediating variables) - occur between two other variables in causal chain. A->B->C. B is intervening.
|
|
Moderator variables
|
show how some relationships changes across the levels of an additional variable.
|
|
Experimental Research
|
Purpose - determine and demonstrate cause and effect relationships.
Defining characteristic - active manipulation of an independent variable Strongest experimental designs have random assignment (produces "equivalent" groups) |
|
Logic of Experiment
|
First, form similar groups at start (random assignment if possible).
Second, pretest participants on DV. Third, manipulate IV. Fourth, posttest participants on DV. For example: give pill to experimental group and placebo to control group; see who improves. |
|
Nonexperimental Research
|
By definition, nonexperimental has no manipulation of IV.
Therefore, nonexperimental is not as good as experimental for studying cause and effect. Sometimes nonexperimental categorized as causal-comparative (IV is categorical and DV is quantitative) and correlational (IV and DV quantitative). |
|
“simple case” of causal - comparative research
|
one categorical IV and one quantitative DV.
Example: Gender (IV) and class performance (DV). |
|
“simple case” of correlational research
|
one quantitative IV and one quantitative DV.
Example: Self-esteem (IV) and class performance (DV). |
|
Correlation Coefficient
|
Varies between –1 and +1, and 0 stands for no relationship.
Two characteristics: strength and direction. Strength: the farther from 0, the stronger the relationship. +1 and -1 are strongest. Direction: if positive sign (+.65) positive correlation (two variables move in the same directions). If negative sign (e.g., -.71) negative correlation (two variables move in opposite directions). |
|
Three required conditions for causality
|
relationship, temporal order, and lack of alternative explanation.
|
|
Phenomenology
|
attempts to understand how one or more individuals experience a phenomenon.
Example: interview 20 widows who describe experiences of death of husband. |
|
Ethnography
|
discovers and describes the cultural characteristics of group.
|
|
culture
|
shared attitudes, values, norms, practices, language, and material things of a group of people.
|
|
Case study research
|
provides detailed account of one or more cases.
Example: study classrooms given new curriculum for technology use. |
|
Grounded theory
|
inductively generates and develops a theory from data.
Example: collect data from parents who have pulled children from public schools and develop theory to explain how and why. |
|
Historical research
|
studies people and events from the past.
|
|
Fundamental principle of mixed research
|
mix quantitative and qualitative methods, procedures, and paradigm characteristics to design with complementary strengths and nonoverlapping weaknesses.
|
|
Hypothesis
|
formal statement of the predicted relationship among the variables being investigated.
|
|
Meta-analysis
|
statistical method summarizing results of many studies
|
|
Sampling
|
drawing sample from population
|
|
representative sample
|
similar to the population
|
|
Population
|
full set from which sample is selected.
|
|
Sample
|
set of elements taken from population.
|
|
Statistic
|
numerical characteristic of a sample.
|
|
Parameter
|
numerical characteristic of population.
|
|
Sampling error
|
difference between the value of sample statistic and true value of population parameter.
|
|
Response rate
|
percentage of people in sample who participate in study.
|
|
Sampling frame
|
list of all people (elements) in population.
|
|
Random sampling
|
produces representative samples.
|
|
Nonrandom sampling
|
does not produce representative samples.
|
|
simple random sampling (SRS)
|
SRS is an equal probability sampling method, everyone in sampling frame has equal chance of being in final sample.
|
|
Systematic sampling
|
Also an equal probability sampling method.Three steps:
Determine sampling interval “k” (population size divided by sample size). Randomly select number between 1 and k, and include in sample. Select each kth element. Example: if k is 10 and randomly selected start number between 1 and 10 is 5, then select persons 5, 15, 25, 35, 45, etc. |
|
periodicity
|
cyclical pattern in sampling frame
|
|
Stratified random sampling
|
random sampling within strata and combining cases into overall sample. Take random sample from each group (females, males).
Combine sets of selected people into final sample. |
|
proportional stratified sampling
|
Make sure sizes of subsamples (e.g., males and females) are proportional to their sizes in population.
Proportional stratified sampling is EPSEM |
|
disproportional stratified sampling
|
Subsamples are not proportional to their sizes in the population.
|
|
clusters
|
has more than one unit in it. (Examples: schools, classrooms, and teams.)
|
|
One-stage cluster sampling
|
Randomly select sample of clusters.
Include all individuals in clusters in sample. Example: randomly select 15 classrooms and include all students in the 15 classrooms in sample. |
|
Two-stage cluster sampling
|
First stage: randomly select sample of clusters.
Second stage: take random sample of individuals from each cluster, and include in final sample. Example: randomly select 30 classrooms and randomly select 10 students from each of the 15 classrooms. |
|
Convenience sampling
|
get the most available people for sample
|
|
Quota sampling
|
set quotas or numbers of kinds of people and meet quotas
|
|
Purposive sampling
|
specify type of people needed and locate some who will participate
|
|
Snowball sampling
|
each participant identifies other potential participants who have certain characteristic
|
|
Random selection
|
select sample from population a random sampling technique.
|
|
Random assignment
|
start with group of people and randomly divide into two or more groups.
|
|
narrow confidence interval
|
precise estimates of population characteristics
|
|
Maximum variation sampling
|
select a wide range of cases
|
|
Homogeneous sample selection
|
select homogeneous case or set of cases for intensive study
|
|
Extreme case sampling
|
select cases that represent extremes on dimension
|
|
Typical-case sampling
|
select typical or average cases
|
|
Critical-case sampling
|
select cases known to be important
|
|
Negative-case sampling
|
be on lookout and select cases that might disconfirm your theory
|
|
Opportunistic sampling
|
select useful cases as opportunity arises
|
|
Mixed purposeful sampling
|
mix the sampling strategies just discussed to fit your needs
|
|
Time orientation
|
Do the quantitative and qualitative phases occur concurrently or sequentially
|
|
Concurrent time orientation
|
data collected for quantitative and qualitative phases at approximately same time.
|
|
Sequential time orientation
|
data obtained in stages
|
|
Identical sample relation
|
same people participate in quantitative and qualitative phases of study.
|
|
Parallel sample relation
|
separate quantitative and qualitative samples drawn from the population.
|
|
Nested sample relation
|
participants selected for one phase are subset of participants selected for other phase.
|
|
Multilevel sample relation
|
quantitative and qualitative samples selected from different levels of a population.
|
|
Independent Variable Manipulation
|
Ways to manipulate IV:
Presence or absence (treatment vs. control). Amount (low, medium, high). Type (strategy1, strategy2). |
|
Experimental control
|
eliminating differential influence of extraneous variables.
|
|
Differential influence
|
effect of extraneous variable varies across comparison groups.
|
|
Control
|
refers to achieving constancy across all comparison groups
|
|
Counterbalancing
|
Only used in repeated measures designs (all participants get all conditions, but in different orders).
|
|
Analysis of Covariance
|
statistical technique to match groups.
|
|
Order effects
|
different orders of conditions affect responses.
|
|
Carryover effects
|
one condition continues or lingers into next condition.
|
|
One-group posttest-only design
|
One group exposed to experimental treatment and then measured on DV.
|
|
One-group Pretest-posttest Design
|
Posttest participants after they have been pretested and administered treatment.
|
|
Posttest-only Design with Nonequivalent Groups
|
Group that receives treatment compared with group that did not receive treatment
|
|
Pretest-posttest control-group design
|
Participants randomly assigned to experimental and control groups; both groups pretested on DV; treatment condition given to experimental group only; both groups posttested on DV.
|
|
Posttest-only Control-group Design
|
Participants are randomly assigned to experimental and control groups and posttested after treatment is administered to experimental group.
|
|
Factorial Design
|
Two or more IVs, at least one of which is manipulated, simultaneously studied to determine their independent and interactive effects
|
|
Quasi-Experimental Research Designs
|
an experimental design that does not provide for full control of potential confounding variables. Does not have random assignment.
Makes ruling out rival hypotheses difficult. Makes it difficult to make causal inference |
|
Nonequivalent Comparison Group Design
|
Design that compares results of experimental and control groups after
Control and experimental group are given pretest and Treatment is administered to experimental group and Both groups are posttested. |
|
Interrupted Time-Series Design
|
Single group of participants repeatedly pretested (baseline), administered treatment, and then repeatedly posttested.
|
|
Regression-Discontinuity Design
|
Treatment effect identified by discontinuity in regression line between individuals who score above and below predetermined cutoff score
Used to determine if some special treatment had an effect |
|
Experimental reliability
|
repeatability of the results of a study.
|
|
Experimental validity
|
correctness of the inference made from the study results.
|
|
Internal Validity
|
accuracy of inference that two variables are causally related.
|
|
Ambiguous temporal precedence
|
Threat to internal validity
Inability to specify which variable is causal. Exists in nonexperimental research studies. |
|
History
|
Threat to internal validity Event other than treatment affects DV.
Can exist in one-group pretest-posttest design |
|
Maturation
|
Threat to internal validity Physical or mental changes occurring over time that influence DV.
Examples: age, learning, boredom, fatigue. |
|
Testing
|
Threat to internal validity Changes in scores on posttest result of having taken pretest.
|
|
Instrumentation
|
Threat to interal validity Changes occur in measurement instrument.
Pretest and posttest different. Person collecting data becomes more skilled on second or subsequent measurement. |
|
Regression artifact
|
Threat to internal validity Tendency for extreme scores to regress toward the mean on a second assessment.
Occurs because chance factors contributed to extreme scores. |
|
Differential selection
|
threat to internal validity When participants forming comparison groups have different characteristics.
|
|
Additive and interactive effects.
|
Threat to internal validity Bias resulting from combination of two or more basic threats,
|
|
External Validity
|
Ability to generalize across
Different people in single population. Different populations of people. Different settings. Different times. Different outcomes. Different treatment outcomes. |
|
Population validity
|
ability to generalize from sample to target population and across its subpopulations.
|
|
Ecological validity
|
generalizing across settings.
|
|
Temporal validity
|
generalizing across time
|
|
Treatment variation validity
|
generalizing across variation in treatments.
|
|
Outcome validity
|
generalizing across different but related DVs.
|
|
Construct Validity
|
extent to which higher order construct is accurately represented in study.
|
|
Threats to Construct Validity
|
Treatment diffusion – participants in one treatment condition exposed to some of other treatment condition.
Many others exist. |
|
Statistical Conclusion Validity
|
Correctness of inference that independent and dependent variables are related.
|
|
Descriptive Statistics
|
application of statistical techniques to summarize and make sense of a particular set of data.
|
|
Data set
|
A set of data with the cases in rows and variables in columns
|
|
Frequency Distributions
|
frequency of each unique data value are shown
|
|
Grouped frequency distribution
|
data are grouped into intervals and frequency of each interval shown
|
|
Bar graph
|
uses vertical bars to represent data.
Height of bars shows frequencies of categories. Used for categorical variables. |
|
Histogram
|
graphic showing distribution of quantitative variable.
Looks like bar graph except there is no space between the bars. |
|
Line graph
|
use line(s) to depict information about variable(s).
Simple line graph can show trend. |
|
Scatter plot
|
depict relationship between two quantitative variables.
IV or predictor variable placed on X axis (horizontal axis) and DV on Y axis (vertical axis). |
|
Measures of central tendency
|
a numerical value is obtained that is considered typical of the quantitative variable.
|
|
skewness
|
If normally distributed, no skew skewed to left negatively skewed skewed to right positively skewed
Rule One. If mean less than median, data are skewed to the left. Rule Two. If mean greater than median, data are skewed to the right. |
|
Variance
|
average deviation from the mean (in squared units
|
|
Standard deviation
|
square root of the variance (converts squared units to regular units).
|
|
Normal Curve
|
Has a bell shape.
If data normally distributed then “68, 95, 99.7 percent rule" applies. |
|
Measures of relative standing
|
provide information about a particular score in relation to other scores.
Commonly used measures: percentile ranks and z scores. |
|
Percentile rank
|
percentage of scores in reference group falling below particular score
|
|
z score
|
shows how many standard deviations (SD) raw score falls from mean
|
|
Contingency Tables
|
Displays information in cells formed by the intersection of two or more categorical variables
|
|
Regression analysis
|
used to explain or predict values of quantitative dependent variable based on values of one or more independent or predictor variables.
|
|
Simple regression
|
one quantitative DV and one IV.
|
|
Multiple regression
|
one quantitative DV and two or more IVs
|
|
Regression equation
|
defines regression line
|
|
Regression coefficient
|
predicted change in DV given one unit change in IV, controlling for the other IVs in equation.
|
|
nonexperimental research
|
lack of manipulation of independent variable. Researcher studies what naturally occurs or has already occurred
|
|
post hoc fallacy
|
arguing, after the fact, A must have caused B because you observed in past that A preceded B
|
|
Independent Variables in Nonexperimental Research
|
Categorical IVs that cannot be manipulated: gender, parenting style, learning style, ethnicity, retention in grade, personality type, drug use.
Quantitative IVs that cannot be manipulated: intelligence, age, GPA, personality trait operationalized as quantitative (e.g., level of self-esteem). |
|
Simple case of causal-comparative
|
- one categorical IV (gender) and one quantitative DV (e.g., performance on a math test).
|
|
Simple case of correlational
|
one quantitative IV (level of motivation) and one quantitative DV (performance on math test).
Check correlation coeffiecient. Is observed correlation statistically significant (not due to chance)? Correlation coefficient detects linear (not curvilinear) relationships. |
|
Applying Required Conditions for Causation in Nonexperimental
|
Condition 1: observe relationship.
Difficult to establish conditions 2 and 3 (especially 3). Condition 2: use logic and theory (biological sex occurs before achievement) and design approaches (longitudinal). Condition 3 is serious problem in nonexperimental research Relationship might be "spurious" (non-causal; due to confounding extraneous variable). Condition 3: use logic and theory (list extraneous variables and measure), control techniques (statistical control, matching), and design approaches. |
|
Partial correlation
|
correlation between two quantitative variables after controlling for extraneous variable.
|
|
Cross-sectional
|
data collected at single point in time,
|
|
Longitudinal or prospective
|
data collected in forward direction at two or more time points (moving forward)
|
|
Retrospective
|
data collected looking backward or from past.
|
|
Descriptive (nonexperimental research object dimension)
|
provide picture of status or characteristics of situation or phenomenon (teachers personality types on Myers-Briggs test).
|
|
predictive (nonexperimental research object dimension)
|
predict future status on dependent variable (predict dropping out of school).
|
|
Explanatory (nonexperimental research object dimension)
|
explain how and why phenomenon operates; interest is in cause-and-effect (test causal model of dropping out).
|
|
Causal modeling
|
constructing theoretical models and testing with new data. Commonly used in nonexperimental research.
|
|
Inferential statistics
|
inferences about characteristics of populations based on sample data.
Goal: go beyond sample data; make inferences about population parameters. |
|
Parameters
|
numerical characteristics of populations.
Examples: population mean, population correlation. Symbolized with Greek letters. |
|
Statistics
|
numerical characteristics of samples.
Examples: sample mean, sample correlations. Symbolized with English letters. Greek letter mu (i.e., µ) symbolizes population mean Roman/English letter X with a bar over it, (called X bar), symbolizes sample mean. |
|
Sampling Distributions
|
Sampling distributions allows researcher make "probability" statements in inferential statistics.
|
|
standard error
|
Standard deviation of sampling distribution
|
|
Point estimate
|
single number. value of sample statistic (sample mean, sample correlation) used to estimate value of population parameter (population mean, population correlation).
|
|
Interval estimate or confidence interval
|
range of numbers range of numbers inferred from sample that has known probability of capturing population parameter over long run (over repeated sampling).
|
|
Why not always use 99% rather than 95% intervals?
|
Answer: 99% interval has to be wider (less precise) than 95%.
Larger sample sizes produce more narrow confidence intervals. Lower levels of confidence (95% rather than 99%) produce more narrow confidence intervals. |
|
Hypothesis testing
|
used to determine when null hypothesis can be rejected in favor of alternative hypothesis.
|
|
Null hypothesis
|
usually prediction of no relationship in population. It states population value assumed for purpose of statistical testing.
|
|
Alternative hypothesis
|
logical opposite of null; says there is relationship in population.
Researchers hope to “nullify” the null and accept alternative hypothesis. |
|
When do I reject the null hypothesis of no relationship and make decision to tentatively accept alternative hypothesis?
|
Reject the null when probability of your result (assuming the null is true) is very small.
Fail to reject the null when probability of your result not small (i.e., your result is not a rare event when the null is true). |
|
significance level
|
the point at which you would consider a result to be very unlikely. Usually .05
|
|
Type I error (hypothesis testing)
|
“false positive”; rejecting a true null; making claim of relationship when there is none.
|
|
Type II error (hypothesis testing)
|
“false negative”; failure to reject false null; making claim of no relationship when there is a relationship in the population.
|
|
t Test for Independent Samples
|
Determine if difference between two group means statistically significant.
|
|
One-Way Analysis of Variance
|
Compare two or more group means for statistical significance.
|
|
Post Hoc Tests in Analysis of Variance
|
Needed when ANOVA used to test three or more means to determine which means are significantly different.
If ANOVA used when just two means, no need for post hoc tests. |
|
The t Test for Correlation Coefficients
|
Determine whether observed correlation coefficient statistically significant.
|
|
The t Test for Regression Coefficients
|
Determine whether regression coefficient is statistically significant.
|
|
The Chi-Square Test for Contingency Tables
|
Determine whether relationship in contingency table statistically significant.
|
|
Modernism
|
used by postmodernists for outdated period in science that viewed the world as static (i.e., unchanging) machine.
|
|
Positivism
|
used by qualitative researchers for “scientism” (true knowledge must be based on science).
|
|
Postmodernism
|
- movement in opposition to modernism; it emphasizes primacy of individuality, difference, fragmentation, flux, constant change, lack of foundations, and interpretation.
|
|
Structuralism
|
emphasizes importance of cultural-structural-institutional and functional relations in language and society; “structure” influences humans’ thinking and behavior.
|
|
Poststructuralism
|
rejects parts of structuralism, but also builds on it.
|
|
Poststructuralism
|
rejects universal truth; emphasizes differences, deconstruction, interpretation, and power of “knowledge.”
|
|
Ethnography
|
discovery and description of culture of group of people.
|
|
Ethnology
|
comparative study of cultural groups
|
|
Ethnohistory
|
study of cultural past of group
|
|
Ethnocentrism
|
judging others based on your cultural standards
|
|
Emic perspective/terms
|
specialized words used by insiders in group
|
|
Etic perspective/terms
|
specialized words used by outsiders and social scientists
|
|
Going native
|
identifying completely with group; unable to be objective
|
|
Holism
|
whole is greater than sum of parts; view group as whole unit
|
|
Case
|
bounded system (person, a group, activity, process)
|
|
Intrinsic case study
|
interest in understanding particulars of case
|
|
Instrumental case study
|
interest in understanding something more general than case
|
|
Collective case study
|
interest in comparing multiple cases
|
|
Open coding
|
read transcripts line-by- line; identify and code concepts in data
|
|
Axial coding
|
organize concepts and make more abstract
|
|
Selective coding
|
focus on main ideas; develop story; finalize theory
|
|
theoretical saturation
|
no new concepts emerge and theory validated
|
|
Interim analysis
|
cyclical process of collecting and analyzing data.
|
|
memoing
|
recording reflective thoughts and insights
|
|
Three approaches to visual data analysis
|
photo interviewing
Semiotics is study of signs Visual content analysis |
|
photo interviewing
|
Researchers show images to research participants during formal or informal interviews. In photo interviewing analysis, analysis done by participants who examine and “analyze” visual images.
|
|
Semiotics is study of signs
|
People’s clothes, nonverbal gestures, myths, stories. In semiotic visual analysis, researcher identifies and interprets symbolic meanings of visual data.
|
|
Visual content analysis
|
identification and counting of events, characteristics, or other phenomena in visual data; it’s more quantitative than previous two approaches to visual data analysis.
|
|
segmenting
|
Divide data into meaningful analytical units
|
|
Coding
|
marking segments of data with symbols, descriptive words, or category names.
|
|
master list
|
list of all codes used
|
|
Intercoder reliability
|
consistency across different coders.
|
|
Intracoder reliability
|
consistency within single coder.
|
|
Inductive codes
|
developed by researcher by directly examining the data (very popular in QDA).
|
|
A priori codes
|
brought to research study or developed before examining data.
|
|
Co-occurring codes
|
partially or completely overlapping codes; same lines or segments can have more than one code.
|
|
Facesheet codes
|
apply to entire document or case
|
|
Enumeration
|
quantifying data (also called quantitizing).
|
|
hierarchies
|
One important kind of relationship is organizing codes or categories into levels
|
|
Typologies
|
(and taxonomies) are example of Spradley's "strict inclusion
|
|
Diagramming
|
making a sketch, drawing, or outline to show how something works or clarify relationship between parts of a whole.
|
|
Matrix
|
rectangular array formed into rows and columns.
|
|
Boolean operators
|
used to create logical combinations such as AND, OR, NOT, IF, THEN, and EXCEPT.
|
|
Validity in Qualitative research
|
Refers to plausible, credible, or trustworthy research
|
|
Reflexivity
|
self-reflection by researcher about his or her biases
|
|
Descriptive validity
|
Factual accuracy of researcher’s account.
Strategy to obtain descriptive validity: Investigator triangulation. use of multiple investigators to collect and interpret data |
|
Interpretive validity
|
Accurately portraying meaning attached by participants to phenomena.
Strategies used. Participant feedback. Use of low-inference descriptors. |
|
Theoretical Validity
|
Degree to which theoretical explanation fits the data.
Strategies used: Extended fieldwork. Theory triangulation. Pattern matching. Peer review. |
|
Inside-outside validity
|
use of participants’ subjective insider or “native” views and researcher’s “objective outsider” view.
|
|
Paradigmatic validity
|
researcher documents his or her philosophical beliefs about research.
|
|
Commensurability validity
|
researcher makes Gestalt switches between qualitative and quantitative creates “integrate viewpoint.”
|
|
Weakness minimization validity
|
researcher combines qualitative and quantitative to have nonoverlapping weaknesses.
|
|
Sequential validity
|
researcher addresses effects from ordering of qualitative and quantitative phases.
|
|
Conversion validity
|
accuracy of quantitizing and qualitizing data.
|
|
Sample integration validity
|
making appropriate generalizations from mixed samples.
|
|
Political validity
|
carefully addressing interests , values, and viewpoints of multiple stakeholders.
|
|
Multiple validities
|
– degree to which all pertinent validities (quantitative, qualitative, and mixed) are addressed and resolved successfully.
|