3: Measures Flashcards
operational (or operative) definition
The phenomena to
be measured should be defined in terms of the operations
used to measure them. Essential in experimental research*
- In other words, a behaviour or process (cognitive or physiological) is
defined by the operations that are carried out in order to measure it
A good operational definition is: Reliable and Reproducible
ostensive definitions
The phenomena to be observed
should be carefully described (textually, graphically,
photographically, etc.) and examples can be given.
Common in observational research (it is the basis of
ethograms and systematic observations)
what are some examples of construct validity?(conventional meaning lost)
Bruxism, Alcoholism, PMS, ADHD (sub-types),
aggressiveness/aggression
Scales of measurement when spacing between values is not known
- Nominal scales and ordinal scales
scales of measurement when spacing is known
Interval scales and ratio scales
nominal scales
a scale that labels variables into distinct classifications and doesn’t involve a quantitative value or order
categories, taxonomies, typologies.
* Examples: Male/female;
Liberal/ Conservative/
Democrat; ADD+H/ADD-H;
extroverted / introverted.
* Types of statements: x is
different from y;
assignment of labels
ordinal scales
continuum or spectrum of observations. Ranking within a category is possible; often different names AND certainly different quantities. Absolute values are not known.
* Examples: Low, moderate,
high self-esteem; Ranking
of a race.
* Types of statements: x is
greater than y; assignment
of values
interval scales
has no absolute zero point, i.e., arbitrary
* In an interval scale, you have a constant unit and it satisfies the condition 2-1 = 3-2 = n- (n-1), and here there is an additive constant of the form y=ax+b (b can be zero or an other value).
* Common in psychology and neuroscience: rating scales, from 0 to 10 (0 does not mean no liking at all), IQ scale (0 does not mean no intelligence).
ratio scales
has an absolute zero point (the absence of the quantity can be indicated). Zero means zero, i.e., nothing.
* In a ratio scale, the linear
transformation of values
must follow the form
y=ax (b is zero, in fact,
must be zero). The Kelvin
scale of temperatures
follows this model. In
other words, with a ratio
scale, all the operations of
arithmetic may be
performed, with the
numerical values
representing absolute
values of the terms.
* Common in psychology
and neuroscience: from
no response to n degree
of response. Example:
score on a memory test.
implications of scales of measurement
- Nominal and ordinal data can be analysed with a number of non-parametric statistical analyses.
- Interval and ratio scales: Analyzed (typically) with parametric statistical analyses.
- Interval scale: spacing between values is known, BUT:
- a score of 120 is not twice
as more as a score of 60
(e.g., two IQ scores).
- a score of 120 is not twice
- Ratio scale: one value is twice as much as another or no quantity of that variable can exist, i.e.:
- a score of 120 is twice as
more as a score of 60 (e.g.,
two scores on a memory
test with 200 items).
- a score of 120 is twice as
- The four scales represent a hierarchy of information yielded (mnemonic: anagram “NOIR”), from little, to substantial.
what does the mnemonic NOIR mean?
- Nominal: Purely qualitative information.
- Ordinal: Qualitative with some crude degree of “quantity” (items can be ranked).
- Interval: Quantitative; we know by “how much” the values differ.
- Ratio: Quantitative; we know how much of the quantity exists.
what are the basic types of variables
- Independent (“X”)
- Dependent (“Y”)
- Independent vs. dependent: To identify the variables, ask the question “What is the effect of [IV] on [DV]?”
what are independent variables?
Treatment, condition, intervention, factor (e.g., conditions: methods of teaching). Independent variables have “levels” (e.g., 2 levels, low and high doses).
Also called “experimental variable”, “manipulated variable”.
what is the dependent variable?
Outcome, response, result, measure (what you measure). The dependent variable is observed and measured.
independent variables- quantitative and qualitative
- Independent variables: manipulated variables (with levels of treatment conditions).
- Quantitative: Treatments differ in frequency, amount (e.g., dose), degree, etc.
- Qualitative: Treatments differ in kind
dependent variables- quantitative and qualitative
- Dependent variables: measurable response.
- Quantitative: Usually the case, e.g., a score or duration.
- Qualitative: Special procedures needed, e.g., form of treatment.
- Subject variables / classification or categorization variables / individual-difference variables / grouping variables: not independent variables per se. E.g., sex, age, ethnic background, linguistic background, etc.
Qualitative variables- unordered and ordered
- Qualitative: Represent an attribute and can be assigned a unique category. Used to categorize information.
- Unordered: Cannot be ranked or ordered; mutually exclusive categories. Example: Dead or alive.
- Ordered: Categories can be placed in rank order. Example: Very young, young, adult, geriatric
quantitative variables- discrete and continuous
- Quantitative: Values determined by “counting” or numerical measurements.
- Discrete: Only whole number values. Examples: Number of offspring, number of cells, number of heart attacks.
- Continuous: Infinite number of whole and fractional values. Example: Body weight (60 kg, 105.9 kg, etc.).
types of independent variables
environmental or situational variables
instructional variables
subject, participation, or individual differences variables
environmental or situational variables
Variables referring to
the manipulation of the environment, i.e., treatment, tasks, etc
instructional variables
Variables referring to what is told to a participant, instructed, or simply suggested. Applies mostly to human research, but potentially to animal research
Subject, participant or individual differences variables (as quasi-IV’s)
Of variable importance in idiographic versus nomothetic research.
other types of variables
controlled, extraneous, confounding or confounded
controlled variables
Any variable that is controlled or held constant across all treatment conditions of an experiment
extraneous variables
“Threatening variable” or “obscuring factors”; impact on the dependent variable
problems and solutions for extraneous/confounding variables
extraneous/confounded variables are a potential source of variability (experimenter-expectancy effects)
Solutions
* Random assignment of participants to treatment conditions.
* Keeping the extraneous variable constant.
* Matching participants on the extraneous variable.
* Building the extraneous variable into the study or “blocking” (creating blocks); The extraneous variable becomes an independent variable like the others.
* Statistical control of the extraneous variable: ANCOVA. The extraneous variable becomes a covariate
confounding or confounded variable
“confounds”, an extraneous variable (usually unmonitored) that can inadvertently affect another experimental variable
threats to inference
- Experimenter-expectancy effects; confirmation bias
- Demand characteristics: Subject’s own expectancies; can occur in animals
- Subject-predisposition effects: can occur with animals
- Cooperative-subject effect: Want to provide data that will support the research hypothesis
- Screw you effect: Want to sabotage the experiment
- Evaluation apprehension: Want positive evaluation; hypothesis-independent
- Faithful subjects: They just do what there are asked to do. Follow instructions, don’t worry about the hypothesis, don’t try to please
- Placebo effect
goal and techniques for the concept of experimental control
- Goal: Control nuisance variables and distortions
- Techniques:
- Experimental:
- Keep the nuisance
variables constant - Assign subjects randomly
to experimental
conditions»_space;> random
assignment - Design: include the
nuisance variable as one
of the variables
- Keep the nuisance
- Statistical: Statistically
remove the effects of a
nuisance variable»_space;>
analysis of covariance
- Experimental:
types of control groups
Control groups: Do not get the independent variable (manipulation, treatment, etc.) or IV.
* Placebo, Nocebo, Sham
group (e.g., surgery):
Quasi-control groups: Subjects in this group do not get a placebo.
* Sub-type
what are placebos
(“I shall please”):
Control group that receives a “dummy” treatment (e.g., sugar pill) that is assumed to have a positive effect.
what are nocebos
Opposite of placebo»_space;> the effect of the IV is worse, not better.
what are sham groups
e.g surgery
Surgery with no experimental purpose.
quasi-control groups sub-type
Sub-type: Simulator groups; subjects are asked to act as if they had received the treatment. Useful in double-blind experiments to detect and assess experimenter- expectancy effects*
- as well as demand characteristics
basic techniques of the concept of experimental control
single-blind procedures/experiments
double-blind procedures/experiments,
partial-blind procedures/experiments
deception
disguised-experiment technique or unobtrusive experimentation
debriefing
what are single-blind procedures/experiments
Minimized demand characteristics;
not always possible because of informed consent requirements
what are Double-blind procedures/experiments
Will reduce both demand
characteristics and experimenter-expectancy effects
what are partial-blind procedures/experiments
Experimenter is not in the know until just before the treatment is about to be administered. Can minimize experimenter-expectancy effects
advanced techniques in the concept of experimental control
multiple researchers
experimental-expectancy control groups
unrelated experiment technique
what is debriefing
Asking the subjects about what they think happened.
Identifies problems with demand characteristics
what is the multiple researchers techniques
experimenters, observers, handlers, trainers, caretakers; useful if you suspect experimenter effects
what is the experimental-expectancy control groups
3 groups of researchers are formed:
* Group 1: Led to expect one experimental outcome (e.g., increase in heart rate)
* Group 2: Led to expect an other experimental outcome, typically opposite (e.g., decrease in heart rate)
* Group 3: Led to expect that the treatment will have no effect (e.g., no change in heart rate)
what is the unrelated experiment technique
Separate the presentation of the IV over the DV. Two
experiments are necessary (within design: subjects do both).
* Experiment 1: Subjects receive the IV
* Experiment 2: Subjects receive the DV
The subjects are told the two experiments are not connected; but they are THE experiment when combined
types of sampling from a population (random or probabilistic sampling)
- Random sampling: The ideal method
- Simple random sampling
- Stratified sampling: Random sampling within segments (strata) of the population based on specific characteristics are considered, e.g., age, gender, etc.
- Cluster sampling: Naturally occurring units of individuals
(groups) are randomly selected, e.g., children in a day care group, students in this class. - Proportionate sampling (to avoid over-representations)
types of sampling from a population (non-random or non-probabilistic sampling)
- Haphazard sampling (not true random sampling)* (ex: grabbing fish from a tank)
- Convenience sampling
- Volunteer sampling (with humans only; “self-selection”)
- Systematic sampling: e.g., every kth element is sampled
- Sequential sampling: Gradual, one at a time
- Quota sampling: Stratified without randomness, i.e., based on convenience sampling principles (non-random).
- Purposive/selective sampling: Selection based on pre-
determined criterion or criteria.
random sampling (or selection)
Selecting subjects from a
population so all members of the population have the same
chance of being selected for a sample (and all possible
samples have the same chance of being selected)
random assignment (or randomization)
Using random chance to determine which condition a participant will experience.
Random assignment determines the difference between experimental and quasi-experimental research.
what is a heuristic
a general rule that is usually correct
