CHAPTER 4: RELIABILITY, VALIDITY & MEASUREMENT
WHAT IS MEASUREMENT?
● MEASUREMENT: assignment of scores that represent a characteristic for the
individual**
○ we do this all the time, we measure our height & weight. we measure temperature
of food or how many guests are invited to a party. We also measure how long we
want our hair when we get a hair cut (shoulder, length, bob, pixie)
PSYCHOLOGICAL CONSTRUCTS -
● emotions, behavior & thoughts
● PSYCHOMETRICS: measurement for psychologists
● examples: backward digit span task for working memory - helps measure brain capacity,
IQ scores, MMPI (Minnesota Multiphase Personality Inventory - measures whether
someone's a psychopath or not).
● Remember our Scientific Method: we are using a systematic procedure for assigning
scores to measure a construct.
● The types of things psychologists want to measure are usually not straight forwards
● CONSTRUCTS: the things that we need to measure that we cannot observe directly,
but are tendencies, like feelings, thoughts, abilities, & attitudes. These are often
internal processes.
● CONCEPTUAL DEFINITION: Focuses on how constructs are made up & how it
relates to other variables
○ NEUROTICISM - positively correlated w/ negative emotions, is stable over
time, & might be genetic
● Psychologists must:
○ Propose a conceptual definition
○ Test it empirically
○ Revise as necessary
OPERATIONAL DEFINITIONS:
● DEFINITiON: How a variable will be measured
● Self-report is one of the most accessible means of gathering data
○ however can be very biased as people may want to appear themselves in a good
light
○ or people may not know themselves that well.
, ● Behavioral measures
● Physiological measures - measures your body reaction (heart rate, etc.)
● Convergent Operation: when you can capture all three
○ allows us to draw broad conclusions instead of narrow ones based on
measurement
LEVELS OF MEASUREMENTS:
1. NOMINAL
1. lowest level of measurement for categorical variables
2. martial status or ethnicity; no order just categories
3. gives you the mode (what number is being repeated)
2. ORDINAL
1. rankings. there is a meaningful higher or lower such as: Satisfied, Very Satisfied,
& completely satisfied.
2. gives you the median or mode (median means the number in the middle)
3. INTERVAL
1. Numerical w/ same intervals: like temperatures (degree) or IQ, But no meaningful
zero - THERE IS ALWAYS A MEASUREMENT SO THERES NO 0
4. RATIO
1. All of the above w/ a true “zero” that means the absence of the construct. Like if
you can someone how many kids they have or how much money they have.
2. give the media, mean or mode
TYPES OF RELIABILITY
TYPE DEFINITION EXAMPLE
Face Validity Face validity refers to Items assessing trust usually
whether a test or look like they are measuring
measurement appears to trust
measure what it’s supposed
to, at face value. In simple
terms, it’s about whether the
test looks like it makes sense
for what it’s trying to assess.
For example, if a math test
has math problems, it has
good face validity because it
seems to measure math skills.
WHAT IS MEASUREMENT?
● MEASUREMENT: assignment of scores that represent a characteristic for the
individual**
○ we do this all the time, we measure our height & weight. we measure temperature
of food or how many guests are invited to a party. We also measure how long we
want our hair when we get a hair cut (shoulder, length, bob, pixie)
PSYCHOLOGICAL CONSTRUCTS -
● emotions, behavior & thoughts
● PSYCHOMETRICS: measurement for psychologists
● examples: backward digit span task for working memory - helps measure brain capacity,
IQ scores, MMPI (Minnesota Multiphase Personality Inventory - measures whether
someone's a psychopath or not).
● Remember our Scientific Method: we are using a systematic procedure for assigning
scores to measure a construct.
● The types of things psychologists want to measure are usually not straight forwards
● CONSTRUCTS: the things that we need to measure that we cannot observe directly,
but are tendencies, like feelings, thoughts, abilities, & attitudes. These are often
internal processes.
● CONCEPTUAL DEFINITION: Focuses on how constructs are made up & how it
relates to other variables
○ NEUROTICISM - positively correlated w/ negative emotions, is stable over
time, & might be genetic
● Psychologists must:
○ Propose a conceptual definition
○ Test it empirically
○ Revise as necessary
OPERATIONAL DEFINITIONS:
● DEFINITiON: How a variable will be measured
● Self-report is one of the most accessible means of gathering data
○ however can be very biased as people may want to appear themselves in a good
light
○ or people may not know themselves that well.
, ● Behavioral measures
● Physiological measures - measures your body reaction (heart rate, etc.)
● Convergent Operation: when you can capture all three
○ allows us to draw broad conclusions instead of narrow ones based on
measurement
LEVELS OF MEASUREMENTS:
1. NOMINAL
1. lowest level of measurement for categorical variables
2. martial status or ethnicity; no order just categories
3. gives you the mode (what number is being repeated)
2. ORDINAL
1. rankings. there is a meaningful higher or lower such as: Satisfied, Very Satisfied,
& completely satisfied.
2. gives you the median or mode (median means the number in the middle)
3. INTERVAL
1. Numerical w/ same intervals: like temperatures (degree) or IQ, But no meaningful
zero - THERE IS ALWAYS A MEASUREMENT SO THERES NO 0
4. RATIO
1. All of the above w/ a true “zero” that means the absence of the construct. Like if
you can someone how many kids they have or how much money they have.
2. give the media, mean or mode
TYPES OF RELIABILITY
TYPE DEFINITION EXAMPLE
Face Validity Face validity refers to Items assessing trust usually
whether a test or look like they are measuring
measurement appears to trust
measure what it’s supposed
to, at face value. In simple
terms, it’s about whether the
test looks like it makes sense
for what it’s trying to assess.
For example, if a math test
has math problems, it has
good face validity because it
seems to measure math skills.
