wobbles, humps and sudden jumps 1 A theoretical reflection on the nature of psychological properties …
Dec 17, 2015
wobbles, humps and sudden jumps 1
A theoretical reflection on the nature of psychological properties …
A theoretical reflection on the nature of psychological properties …
wobbles, humps and sudden jumps - theoretical reflection 2
Dynamic view of psychological phenomenaDynamic view of psychological phenomena
• The principal focus is on the individual person• For whom actions, thoughts, concepts, skills,
…• Are the product of a real-time construction
process• that requires organismic and contextual
components• that emerges out of the interplay between
organism and context
• The principal focus is on the individual person• For whom actions, thoughts, concepts, skills,
…• Are the product of a real-time construction
process• that requires organismic and contextual
components• that emerges out of the interplay between
organism and context
If this is our framework, …And we wish to measure psychological variables/properties …
We are confronted with a rather serious difficulty...
If this is our framework, …And we wish to measure psychological variables/properties …
We are confronted with a rather serious difficulty...
wobbles, humps and sudden jumps - theoretical reflection 3
Why?Why?
• Measurement = Attribution of a specific property (variable) as a definite property of a person
• But, according to a dynamic/contextualist view• There are no such properties of persons• “there is no object concept in the child, it’s assembled each
time the child is in a context that requires it…”
• Error handling requires Conditional Independence• but
• in an individual process, the next step usually depends on the preceding one: conditional dependence
• The structure of variability across a group structure of variability across an individual process (Molenaar)
• Measurement = Attribution of a specific property (variable) as a definite property of a person
• But, according to a dynamic/contextualist view• There are no such properties of persons• “there is no object concept in the child, it’s assembled each
time the child is in a context that requires it…”
• Error handling requires Conditional Independence• but
• in an individual process, the next step usually depends on the preceding one: conditional dependence
• The structure of variability across a group structure of variability across an individual process (Molenaar)
A dynamic, process-oriented, contextualist view requires a rethinking of psychometric axioms…
What is the nature of a psychological property, how should it be measured and represented?
A dynamic, process-oriented, contextualist view requires a rethinking of psychometric axioms…
What is the nature of a psychological property, how should it be measured and represented?
wobbles, humps and sudden jumps - theoretical reflection 4
The color of her socks… (1 of 2)The color of her socks… (1 of 2)
What is the color of her socks? Yellow…
The color cannot be observed directly (she won’t show you, she’s too prudish)But she will tell you the color by pointing at a spot on a color chart
Possibility of Measurement Error!!
Both the color of the socks she wears and the color spot she points at are determinate properties…
Difference between the spot and the true color = measurement error
Both the color of the socks she wears and the color spot she points at are determinate properties…
Difference between the spot and the true color = measurement error
wobbles, humps and sudden jumps - theoretical reflection 5
The color of her socks… (2 of 2)The color of her socks… (2 of 2)
What is the color of the socks that she will wear tomorrow?
The color has not yet been determined… anything goes?No, we can provide information about tomorrow’s colorBy specifying constraints on the degrees of freedom
•Based on the socks in her drawer, socks that are in the wash, the color she was wearing today, …
The color of the socks she will be wearing tomorrow is an indeterminate property
Which can be described by specifying constraints on degrees of freedom
The color of the socks she will be wearing tomorrow is an indeterminate property
Which can be described by specifying constraints on degrees of freedom
wobbles, humps and sudden jumps - theoretica lreflection 6
Back to psychological properties… (1 of 2)Back to psychological properties… (1 of 2)
• The standard view• Just like the color of the socks she is wearing, but refuses
to show …• Psychological properties are determinate properties (fixed
or true value)• But they are covert, latent, concealed, …• And thus, must be measured indirectly, which implies
measurement error (I.e variability in the measure)• Example: true score theory
• Observed score = true score + error
• Example: “true category” theory• Object Concept, ADHD
• The standard view• Just like the color of the socks she is wearing, but refuses
to show …• Psychological properties are determinate properties (fixed
or true value)• But they are covert, latent, concealed, …• And thus, must be measured indirectly, which implies
measurement error (I.e variability in the measure)• Example: true score theory
• Observed score = true score + error
• Example: “true category” theory• Object Concept, ADHD
wobbles, humps and sudden jumps - theoretical reflection 7
Back to psychological properties… (2 of 2)Back to psychological properties… (2 of 2)
• The contextualist view• What we call psychological properties are
properties that become determinate in a specific context, action, ..
• They fluctuate and vary ..• However, we wish to make statements about
such properties that go beyond an actual situation
• In that sense, psychological properties are indeterminate properties
• They are specified by describing constraints on the degrees of freedom
• The contextualist view• What we call psychological properties are
properties that become determinate in a specific context, action, ..
• They fluctuate and vary ..• However, we wish to make statements about
such properties that go beyond an actual situation
• In that sense, psychological properties are indeterminate properties
• They are specified by describing constraints on the degrees of freedom
wobbles, humps and sudden jumps - theoretical reflection 8
0 63 9 1512 18 21 24
1313
Measurement erro
r
Measurement erro
r
The standard view (1 of 2)The standard view (1 of 2)
We need a ruler ...We are dealing with determinate, but “hidden” properties …
We need a ruler ...We are dealing with determinate, but “hidden” properties …
wobbles, humps and sudden jumps - theoretical reflection 9
0 63 9 1512 18 21 24
The standard view (2 of 2)The standard view (2 of 2)
The score is based on a probability…Which can, in principle, be sampled an infinite number of times …Each sample is, in principle, independent of any other one …
The score is based on a probability…Which can, in principle, be sampled an infinite number of times …Each sample is, in principle, independent of any other one …
1115 13 12 14 13
1312
15The true value of the variable…
Is a hidden but determinate property:The average of all the values that can be sampled from a probability distributionWhich implies (in principle) the possibility of
•infinite sampling•independent sampling
The true value of the variable…
Is a hidden but determinate property:The average of all the values that can be sampled from a probability distributionWhich implies (in principle) the possibility of
•infinite sampling•independent sampling
wobbles, humps and sudden jumps - theoretical reflection 10
An alternative view (1 of 6)An alternative view (1 of 6)
• We need a ruler to specify an indeterminate property
• We learn about this property by measuring a number of determinate cases
• And then use this information to specify the indeterminate property as constraints on the degrees of freedom
• By specifying a characteristic range (instead of a true score)
• We need a ruler to specify an indeterminate property
• We learn about this property by measuring a number of determinate cases
• And then use this information to specify the indeterminate property as constraints on the degrees of freedom
• By specifying a characteristic range (instead of a true score)
wobbles, humps and sudden jumps - theoretical reflection 11
0 63 9 1512 18 21 24
An alternative view (2 of 6)An alternative view (2 of 6)
char
acte
rist
icn
ess Range of
characteristic scores
Range of uncharacteristic scores
Range of uncharacteristic scores
The white trapezium line specifies the characteristicness (0 to 1) of the corresponding scores.
It corresponds with a standard function from Fuzzy Logic, namely the degree-of-membership function.
It specifies the degree-of-membership of all possible scores, to the set of characteristic scores.
The function specifies an open set of scores.
The open set corresponds with the notion of an indeterminate property (properties that have yet to be determined, but are governed by constraints)
The membership (or characteristicness) function•characterizes an individual person•implies both person and context properties (the person’s characteristic contexts)•is not a probability function
The white trapezium line specifies the characteristicness (0 to 1) of the corresponding scores.
It corresponds with a standard function from Fuzzy Logic, namely the degree-of-membership function.
It specifies the degree-of-membership of all possible scores, to the set of characteristic scores.
The function specifies an open set of scores.
The open set corresponds with the notion of an indeterminate property (properties that have yet to be determined, but are governed by constraints)
The membership (or characteristicness) function•characterizes an individual person•implies both person and context properties (the person’s characteristic contexts)•is not a probability function
wobbles, humps and sudden jumps - theoretical reflection 12
0 63 9 1512 18 21 24
An alternative view (3 of 6)An alternative view (3 of 6)
char
acte
rist
icn
ess
Uncharacteristic scoreIf it occurs, the context is prob-ably highly uncharacteristic …Or, the dynamic process that leads to the score is probably anomalous …Likelihood: very low
Characteristic scoreIf it occurs, the context and/or the dynamic process that leads to the score is probably highly characteristic.Likelihood: high
“Not very” characteristic scoreIf it occurs, the context and/or the dynamic process that leads to the score are probably “not very” characteristic.Likelihood: relatively high
The members of the open set of scores are actual scores, i.e actual measurements or observations.
The open set of scores has •A characteristic number of members (measurements, observations)•Which have a characteristic temporal order(note the difference with a probability distribution)
The membership (or characteristicness) function need not be unimodal, but can also be bimodal or n-modal
The property of modality can be used to describe different forms of developmental transitions
The members of the open set of scores are actual scores, i.e actual measurements or observations.
The open set of scores has •A characteristic number of members (measurements, observations)•Which have a characteristic temporal order(note the difference with a probability distribution)
The membership (or characteristicness) function need not be unimodal, but can also be bimodal or n-modal
The property of modality can be used to describe different forms of developmental transitions
wobbles, humps and sudden jumps - theoretical reflection 13
0 63 9 1512 18 21 24
An alternative view (4 of 6)An alternative view (4 of 6)
char
acte
rist
icn
ess
Ranges need not be unimodal …Multimodality: e.g. two sets of constraints and two corresponding ranges
Ranges need not be unimodal …Multimodality: e.g. two sets of constraints and two corresponding ranges
Examples
Fischer: testing with and without support, optimal and functional levelGoldin-Meadow and Alibali: different modes of thought under different forms of expressionVan der Maas and Molenaar: simultaneous modes of conservation
Examples
Fischer: testing with and without support, optimal and functional levelGoldin-Meadow and Alibali: different modes of thought under different forms of expressionVan der Maas and Molenaar: simultaneous modes of conservation
wobbles, humps and sudden jumps - theoretical reflection 14
An alternative view (5 of 6)An alternative view (5 of 6)
• A range is not a probability• infinite sampling, conditional independence
• The dynamics of a variable determine its sampling characteristics• Characteristic sampling order
• Repeated measurement of ToM in normal children: learning effect and increase in score
• In PDD-NOS children, repeated testing causes a significant drop in the scores
• Dynamic testing
• characteristic sampling frequency• Language use <> self and identity discourse
• A range is not a probability• infinite sampling, conditional independence
• The dynamics of a variable determine its sampling characteristics• Characteristic sampling order
• Repeated measurement of ToM in normal children: learning effect and increase in score
• In PDD-NOS children, repeated testing causes a significant drop in the scores
• Dynamic testing
• characteristic sampling frequency• Language use <> self and identity discourse
wobbles, humps and sudden jumps - theoretical reflection 15
If you sample from a range, it will have a characteristic order and frequency
If you sample from a range, it will have a characteristic order and frequency
0 63 9 1512 18 21 24
An alternative view (6 of 6)An alternative view (6 of 6)
char
acte
rist
icn
ess
Taking a test is an action …People learn from their actions (dynamic testing)
Taking a test is an action …People learn from their actions (dynamic testing)
wobbles, humps and sudden jumps - theoretical reflection 16
A toolbox of rulersA toolbox of rulers
• The variable ruler • Specifying both determinate and indeterminate properties• e.g. a test, frequencies
• The characteristicness ruler• Degree of characteristicness• Applies to properties of the person as well as the
environment
• The time line: short-, middle and long-term change• The degree-of-membership ruler
• Diagnostic categories, linguistic categories, …
• The variable ruler • Specifying both determinate and indeterminate properties• e.g. a test, frequencies
• The characteristicness ruler• Degree of characteristicness• Applies to properties of the person as well as the
environment
• The time line: short-, middle and long-term change• The degree-of-membership ruler
• Diagnostic categories, linguistic categories, …