Kwantitatieve analyse van individuele data PAMA5108
College 1 Introduction – Analysing individual development: what, why and
how?
Today
Why do we need to focus on individual development and use these complicated methods?
The problem with inferring individual development from group-based analyses
The Dynamic systems theory perspective: Analyzing changes within individuals
Variability
The problem of the ergodicity assumption
Group to individual generalizibality
In classic developmental research group-level analyses are often used to make
individual level claims
Non-Ergodicity: relations between variables in a group are often different from the
relations within an individual over time
Group level results could be different from individual results. Example typing speed
and accuracy.
Two assumptions for generalizing from group to individual
Stationarity: an individual remains the same over time (mean, variance etc.)
Homogeneity: each individual is the same: for everyone the same relations between
variables.
Hower,These assumptions almost never checked: but unlikely that these assumptions
generally hold true for humans. People change = non stationary and people differ =
non hemogeneity. Huge umplications we cannot use group based findings for
designing interventions.
Examples of non-ergodicity from developmental pychology
Educational commitment of 1st year students over 30 weeks. If you look at the individual
students its differents, so we can’t bring group level to individuals. Group-level is mostly
average effect, but in individuals there is more variability.
Example effect of career choice intervention and commitment
,What we know: Trajectories differ greatly. Individuals likely also differ in what step is best for
them as an intervention.
What we do not know: why not all participants benefit, what works for whom and when, two
measurement points do not tell about the process.
Dynamic systems perspective: modeling individual trajectories
Changes in an indivudal depend on the continuous and non-linear interaction between the
person and the context
Key concepts
Interativity: current state depens on previous state
Self organisation: changes emerge in the whole system and there is no simple
causality attractor states (stable state, hoe dieper hoe moeilijker om er vanaf te
komen, kan zowel negatief als positief zijn)
Changes can be irregular, discontinuous and delayed
Understanding the effects of an intervention is possible only on the basis of
individual trajectories.
In research and practice we need knowledge about: the developmental patterns in one
individual/system. The relations between variables over time within an individual/system.
The effects of interventions in an individual/system.
To grasp the process we need to focus on: individual trajectories, non-lineairity, complex
interactions between variables
How can we describe trajectories?
Variability
Slope
Level (atrractors)
Relations between variables.
Changes in variability
Changes in slope (knicks)
Changes in level (jumps and transitions)
Changes in relations between variables
Different types of variability
Variability can refer to the fluctuations around a person’s own average= standard deviation.
,Variability can refer to the fluctuations from one time-point to the next = MASD (Mean
absolute successive difference), measurement of variability based on the difference between
each time point and the next
Meaning of variability
First meaning: Studied as stable characteristic (thus comparison between individuals, or
between observations over timevery low variability indicates rigidity, higer variability
indicates a more flexible system, too much variability may become chaotic.
Second meaning: Studied as dynamic property thus, as changes in variability during the
process within one person increases in variability may predict transitions, decreases
indicate increase of stability, higher and lower periods of variability may be related to the
context at those moments.
What questions can variability help to answer?
What is the effect of my behaviour change? Does this client improve? How do my pupils
develop during my new training scheme?
There are many different methods to study individual development, but they all used to
study or model individual changes over time.
, Tweede deel college
Vragen gaan over kennis en begrip van de methoden toegepast op het vakgebied, de
theoretische achtergrond, de toepassing, de interpretaties van resultaten, de aannames en
de afwegingen die een rol spelen.
Welke kwantiatieve data?
Van processen (developmental science). Processen zijn per definitie individueel. Individu kan
ook een klas of groep zijn, individuele eenheid of systeem > een echte eenheid, dus geen
gemiddelden (fictief getal)
Ergodiciteit
Alle variabele (taalniveau, rekenniveau, geluk), personen en tijd kun je in een kubus
samenvatten. Ergodiciteit zegt: ‘het maakt niet uit of je het sneetje van de onderkant of van
de zijkant pakt’. Dit maakt wel veel uit, krijg je totaal andere data van.
Bijvoorbeeld leesontwikkeling van kinderen volgen in groep 3. Op groepsniveau (cross-
sectioneel of gemiddeld longitudinaal) zie je een geleidelijke steigende lijn, dit betekent dat
gemiddeld kinderen steeds beter leren lezen. Op individueel zie je dan dat kind A een snelle
sprong maakt en daarna stagneert. Kind B sprongsgewijs en kind C heeft sterke fluctuaties.
Het groepsgemiddelde geeft een glad stijgend patroon, maar geen enkel kind ontwikkelt zich
precies zo.
In een Idiografisch onderzoek start je bij het individu of het systeem, dan beschrijf je de
dynamiek/processen en vervolgens kijk je bottom-up naar overeenkomsten. We zijn gewend
om nomethethisch onderzoek te doen, gericht op het destilleren van algemene principes (die
College 1 Introduction – Analysing individual development: what, why and
how?
Today
Why do we need to focus on individual development and use these complicated methods?
The problem with inferring individual development from group-based analyses
The Dynamic systems theory perspective: Analyzing changes within individuals
Variability
The problem of the ergodicity assumption
Group to individual generalizibality
In classic developmental research group-level analyses are often used to make
individual level claims
Non-Ergodicity: relations between variables in a group are often different from the
relations within an individual over time
Group level results could be different from individual results. Example typing speed
and accuracy.
Two assumptions for generalizing from group to individual
Stationarity: an individual remains the same over time (mean, variance etc.)
Homogeneity: each individual is the same: for everyone the same relations between
variables.
Hower,These assumptions almost never checked: but unlikely that these assumptions
generally hold true for humans. People change = non stationary and people differ =
non hemogeneity. Huge umplications we cannot use group based findings for
designing interventions.
Examples of non-ergodicity from developmental pychology
Educational commitment of 1st year students over 30 weeks. If you look at the individual
students its differents, so we can’t bring group level to individuals. Group-level is mostly
average effect, but in individuals there is more variability.
Example effect of career choice intervention and commitment
,What we know: Trajectories differ greatly. Individuals likely also differ in what step is best for
them as an intervention.
What we do not know: why not all participants benefit, what works for whom and when, two
measurement points do not tell about the process.
Dynamic systems perspective: modeling individual trajectories
Changes in an indivudal depend on the continuous and non-linear interaction between the
person and the context
Key concepts
Interativity: current state depens on previous state
Self organisation: changes emerge in the whole system and there is no simple
causality attractor states (stable state, hoe dieper hoe moeilijker om er vanaf te
komen, kan zowel negatief als positief zijn)
Changes can be irregular, discontinuous and delayed
Understanding the effects of an intervention is possible only on the basis of
individual trajectories.
In research and practice we need knowledge about: the developmental patterns in one
individual/system. The relations between variables over time within an individual/system.
The effects of interventions in an individual/system.
To grasp the process we need to focus on: individual trajectories, non-lineairity, complex
interactions between variables
How can we describe trajectories?
Variability
Slope
Level (atrractors)
Relations between variables.
Changes in variability
Changes in slope (knicks)
Changes in level (jumps and transitions)
Changes in relations between variables
Different types of variability
Variability can refer to the fluctuations around a person’s own average= standard deviation.
,Variability can refer to the fluctuations from one time-point to the next = MASD (Mean
absolute successive difference), measurement of variability based on the difference between
each time point and the next
Meaning of variability
First meaning: Studied as stable characteristic (thus comparison between individuals, or
between observations over timevery low variability indicates rigidity, higer variability
indicates a more flexible system, too much variability may become chaotic.
Second meaning: Studied as dynamic property thus, as changes in variability during the
process within one person increases in variability may predict transitions, decreases
indicate increase of stability, higher and lower periods of variability may be related to the
context at those moments.
What questions can variability help to answer?
What is the effect of my behaviour change? Does this client improve? How do my pupils
develop during my new training scheme?
There are many different methods to study individual development, but they all used to
study or model individual changes over time.
, Tweede deel college
Vragen gaan over kennis en begrip van de methoden toegepast op het vakgebied, de
theoretische achtergrond, de toepassing, de interpretaties van resultaten, de aannames en
de afwegingen die een rol spelen.
Welke kwantiatieve data?
Van processen (developmental science). Processen zijn per definitie individueel. Individu kan
ook een klas of groep zijn, individuele eenheid of systeem > een echte eenheid, dus geen
gemiddelden (fictief getal)
Ergodiciteit
Alle variabele (taalniveau, rekenniveau, geluk), personen en tijd kun je in een kubus
samenvatten. Ergodiciteit zegt: ‘het maakt niet uit of je het sneetje van de onderkant of van
de zijkant pakt’. Dit maakt wel veel uit, krijg je totaal andere data van.
Bijvoorbeeld leesontwikkeling van kinderen volgen in groep 3. Op groepsniveau (cross-
sectioneel of gemiddeld longitudinaal) zie je een geleidelijke steigende lijn, dit betekent dat
gemiddeld kinderen steeds beter leren lezen. Op individueel zie je dan dat kind A een snelle
sprong maakt en daarna stagneert. Kind B sprongsgewijs en kind C heeft sterke fluctuaties.
Het groepsgemiddelde geeft een glad stijgend patroon, maar geen enkel kind ontwikkelt zich
precies zo.
In een Idiografisch onderzoek start je bij het individu of het systeem, dan beschrijf je de
dynamiek/processen en vervolgens kijk je bottom-up naar overeenkomsten. We zijn gewend
om nomethethisch onderzoek te doen, gericht op het destilleren van algemene principes (die
