OPERATIONAL DEFINITIONS

OPERATIONAL DEFINITIONS  Measured using reliability coefficients.

 An operational definition specifies the Test-Retest Reliability
precise meaning of a variable within an  Same individuals take the same test
experiment: twice at different times.
 It defines a variable in terms of Interitem Reliability (Internal
observable operations, procedures, and Consistency)
measurements.  Consistency within a test or
 It is called an operational definition questionnaire.
because it clearly describes the  Methods: Split-Half, Cronbach Alpha.
operations involved in manipulating or
measuring the variables in an Validity
experiment.  Face Validity
An operational definition turns abstract o The extent to which a test
ideas into measurable, observable things. appears to measure what it
claims to measure.
DEFINING NONCONSTRUCT VARIABLES o Judged superficially.
 Variables that are directly observable  Content Validity
 Even observable variables can be o The extent to which a test fully
interpreted differently without clear represents the construct it's
definitions. intended to measure.
 Precise operational definitions  Predictive Validity
ensure clarity, replicability, and o How well a test forecasts future
objectivity. performance or behavior.
Even things you can see need clear rules for o Related to the test's ability to
measuring so results are accurate and predict outcomes.
consistent.  Concurrent Validity
o How well a test correlates with
Nonconstruct = stuff you can see or a current established measure
count of the same construct.
Construct = stuff you can’t see, you have o Testing and comparison are
to measure with tools or tests done at the same time.

DEFINING SCALES OF MEASUREMENT CONSTRUCT VALIDITY

 Measurement scales help researchers  The degree to which a test actually
define variables precisely. measures the theoretical construct it's
 They determine how much information intended to measure.
we get from data and which statistical  The most important and broadest form
tests we can use. of validity.
 Nominal, Ordinal, Interval, Ratio  Involves:
 More precise scales (interval/ratio) ○ Convergent validity: Test
provide more data and allow for correlates with other measures of the
stronger statistical analysis. same construct.
○ Divergent (discriminant) validity:
EVALUATING OPERATIONAL DEFINITIONS Test does not correlate with different or
Reliability unrelated constructs.
 Consistency and dependability of a
measurement or procedure. INTERNAL VALIDITY
 A reliable operational definition yields  Internal validity refers to the degree to
similar results across repeated uses. which we can confidently conclude that
Interrater Reliability the independent variable caused the
 Involves multiple observers rating the changes in the dependent variable.
same response.
OPERATIONAL DEFINITIONS
 When low, the observed effect might  Keep detailed records while conducting
be due to something other than the the experiment
independent variable.  Be specific enough that others don't
 We cannot rule ou t alternative have to guess
explanations.  Include anything that might affect
replication or generalizability
 Explains who participated, what
CLASSIC THREATS TO INTERNAL VALIDITY materials were used, how the study
1. History – Events outside the was conducted, and when/where it
experiment affect results. occurred.
 Example: A student’s test score  Should be detailed enough for another
improves because a school-wide researcher to replicate the study
tutoring program happened at the exactly.
same time.
2. Maturation – Participants naturally
change over time.
 Example: Kids getting older might
improve in reading ability regardless of
a reading program.
3. Testing (Practice Effects) – Taking
the test itself affects scores.
 Example: Students do better on a
second math test just because they
practiced the first one.
4. Instrumentation – Changes in
measurement tools or observers
affect results.
 Example: A scale gets miscalibrated, or
a new teacher grades differently than
the old teacher.
5. Statistical Regression (Regression
to the Mean)
o Extreme scores tend to move
closer to the average upon
retesting.
6. Selection Bias
o Groups differ at the start due to
non-random assignment.
7. Mortality (Attrition)
o Participants drop out unequally
between groups.
8. Selection-Maturation Interaction
o Groups differ at the beginning
and change at different rates
over time.

PLANNING THE METHOD SECTION

 Participants (or Subjects)
 Materials / Stimuli / Instruments
 Procedure
Tip: Present events in chronological
order for clarity