Basic terminologies in Research Methodology

Question	Answer
Independent variable	factor believed to bring about change in dependent variable. CAUSE or TREATMENT
Dependent variable	change that results from the intervention (independent variable) OUTCOME
Research/directional hypothesis	predicts expected relationship between variables
Null hypothesis	states that no relationship exists between the variables. relationship found result of chance or sampling error
Null hypothesis accepted	no significant difference observed
Null hypothesis rejected	Signif difference was observed
Nominal data	two or more mutually exclusive categories. male/female, tall/short
Ordinal data	classifies and ranks in terms of the degree they possess a common characteristic. NOT equal intervals. GPA, MMT.
Interval data	Ranks on predetermined EQUAL INTERVALS. NO true ZERO. IQ, F/C temp
Ratio data	ranks based on EQUAL INTERVALS and TRUE ZERO POINT. most precise level of measurement. Goni, weight, force.
Random sampling	all individuals in popn have equal chance of being chosen
Systematic sampliing	selected from a popn by specified intervals
Stratified sampling	Selected from a popn from identified subgroups (height, weight)
Double blind study	subject & investigator not aware of group assignment
Effect size	size of difference between sample means. Allows for a statistical test to find a difference when one really does exist.
Generalizability	the degree to which findings apply to an entire popn
Gold Standard	instrument with established validity that can be used as a standard fro assessing other instruments
Informed consent includes:	info about general nature of study, any risks, what will be done to minimize risks, possible benefits, ethical disclosure.
Control group	group resembling experiemental group that does not receive treatment
Intervening variable	variable that alters the relationship between IND and DEP variables
Validity	degree to which something accurately measures what it is supposed to
Internal validity	degree observed differences in DEP variable are result of IND variable and not some other variable
External Validity	Degree results are generalizable to popn or to environmental settings outside of study
Face Validity	assumption of validity based on appearance of an instrument, psychometrically unsound
Content validity	degree the instrument measures an intended content area - requires item & sampling validity
Concurrent validity	degree scores on one test are related to scores on another criterion test, usu gold standard
Predictive validity	degree test can predict future performance
Construct validity	degree test measures an intended hypothetical abstract concept ( non-observable behaviors or ideas)
Sampling bias	sampling error - sample of convenience, instead of randome selection
Learning effect	pretest influences scores on post test
experimenter bias	expectations of investigator influence results
Hawthorne effect	subject's knowledge of participation in experiment influences results
Placebo effect	subjects respond to a sham with positive effects
Reliability	Degree a test consistently measures what it is intended to
Interrater reliability	consistency of multiple raters
Intrarater reliabilatiy	Individual's consistency in rating
Test-retest reliability	scores on test are stable or consistent over time, measure of instrument stability.
Split half reliability	degree of agreement when test split in half. measure of internal consistency of an instrument
Sensitivity	correctly IDs TRUE POSITIVE
Specificity	correctly IDs TRUE NEGATIVE
Predictive Value	ability to estimate the likelihood that person will test positive or negative for a condition
Independent variable	factor believed to bring about change in dependent variable. CAUSE or TREATMENT
Dependent variable	change that results from the intervention (independent variable) OUTCOME
Research/directional hypothesis	predicts expected relationship between variables
Null hypothesis	states that no relationship exists between the variables. relationship found result of chance or sampling error
Null hypothesis accepted	no significant difference observed
Null hypothesis rejected	Signif difference was observed
Nominal data	two or more mutually exclusive categories. male/female, tall/short
Ordinal data	classifies and ranks in terms of the degree they possess a common characteristic. NOT equal intervals. GPA, MMT.
Interval data	Ranks on predetermined EQUAL INTERVALS. NO true ZERO. IQ, F/C temp
Ratio data	ranks based on EQUAL INTERVALS and TRUE ZERO POINT. most precise level of measurement. Goni, weight, force.
Random sampling	all individuals in popn have equal chance of being chosen
Systematic sampliing	selected from a popn by specified intervals
Stratified sampling	Selected from a popn from identified subgroups (height, weight)
Double blind study	subject & investigator not aware of group assignment
Effect size	size of difference between sample means. Allows for a statistical test to find a difference when one really does exist.
Generalizability	the degree to which findings apply to an entire popn
Gold Standard	instrument with established validity that can be used as a standard fro assessing other instruments
Informed consent includes:	info about general nature of study, any risks, what will be done to minimize risks, possible benefits, ethical disclosure.
Control group	group resembling experiemental group that does not receive treatment
Intervening variable	variable that alters the relationship between IND and DEP variables
Validity	degree to which something accurately measures what it is supposed to
Internal validity	degree observed differences in DEP variable are result of IND variable and not some other variable
External Validity	Degree results are generalizable to popn or to environmental settings outside of study
Face Validity	assumption of validity based on appearance of an instrument, psychometrically unsound
Content validity	degree the instrument measures an intended content area - requires item & sampling validity
Concurrent validity	degree scores on one test are related to scores on another criterion test, usu gold standard
Predictive validity	degree test can predict future performance
Construct validity	degree test measures an intended hypothetical abstract concept ( non-observable behaviors or ideas)
Sampling bias	sampling error - sample of convenience, instead of randome selection
Learning effect	pretest influences scores on post test
experimenter bias	expectations of investigator influence results
Hawthorne effect	subject's knowledge of participation in experiment influences results
Placebo effect	subjects respond to a sham with positive effects
Reliability	Degree a test consistently measures what it is intended to
Interrater reliability	consistency of multiple raters
Intrarater reliabilatiy	Individual's consistency in rating
Test-retest reliability	scores on test are stable or consistent over time, measure of instrument stability.
Split half reliability	degree of agreement when test split in half. measure of internal consistency of an instrument
Sensitivity	correctly IDs TRUE POSITIVE
Specificity	correctly IDs TRUE NEGATIVE
Predictive Value	ability to estimate the likelihood that person will test positive or negative for a condition