30
1 Causal Rasch Models and Individual Growth Trajectories National Center for the Improvement of Educational Assessment January 18, 2011 A. Jackson Stenner Chairman & CEO, MetaMetrics [email protected]

Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

Embed Size (px)

DESCRIPTION

Causal Rasch Models and Individual Growth Trajectories National Center for the Improvement of Educational Assessment January 18, 2011. Jackson Stenner Chairman & CEO, MetaMetrics [email protected]. - PowerPoint PPT Presentation

Citation preview

Page 1: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

1

Causal Rasch Models and Individual Growth Trajectories

National Center for the Improvement of Educational AssessmentJanuary 18, 2011

A. Jackson Stenner

Chairman & CEO, [email protected]

Page 2: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

2

“Although adopting a probabilistic model for describing responses to an intelligence test, we have taken no sides in a possible argument about responses being ultimately explainable in causal terms.”

(Rasch, 1960, p.90)

Page 3: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

3

Three well researched constructs

Reader ability

Text Complexity

Comprehension

Page 4: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

4

Reader Ability

Temperature

Page 5: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

5

Reading is a process in which information from the text and the knowledge possessed by the reader act together to

produce meaning.

Anderson, R.C., Hiebert, E.H., Scott, J.A., & Wilkinson, I.A.G. (1985) Becoming a nation of readers: The report of the Commission on ReadingUrbana, IL: University of Illinois

Page 6: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

6

An Equation

= Reader Ability

Text ComplexityComprehension -

Conceptual

Statistical

RawScore

=i

e (RA – TC )i

1 + e (RA – TC i )

RA = Reading Ability

TC = Text Calibrations

Page 7: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

7

Each of these thermometers is engineered to use the same

correspondence table

Each of these reading tests is engineered to use the same correspondence table

Page 8: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

8

Correspondence Table: Co and Lexile

Raw Score Co Lexil

eRaw Score Co Lexil

eRaw Score

Co LexileRaw Score

Co Lexile

1 35.6 378L 12 36.8 905L 23 38.0 1116L 34 39.

21331L

2 35.7 509L 13 36.9 926L 24 38.1 1134L 35 39.

31355L

3 35.8 589L 14 37.0 947L 25 38.2 1151L 36 39.

41381L

4 35.9 647L 15 37.1 968L 26 38.3 1170L 37 39.

61409L

5 36.0 695L 16 37.2 987L 27 38.4 1188L 38 39.

71440L

6 36.1 736L 17 37.3 1007L 28 38.6 1207

L 39 39.8

1474L

7 36.2 770L 18 37.4 1025L 29 38.7 1226

L 40 39.9

1513L

8 36.3 801L 19 37.6 1044L 30 38.8 1245

L 41 40.0

1560L

9 36.4 830L 20 37.7 1062L 31 38.9 1265

L 42 40.1

1616L

10 36.6 857L 21 37.8 1080L 32 39.0 1286

L 43 40.2

1697L

11 36.7 881L 22 37.9 1098L 33 39.1 1308

L 44 40.3

1829L

Page 9: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

9

Aspect/Construct Temperature Reader Ability

Object of measurement Person Person

Instrument Thermometer Reading test

Measurement outcome Number of theory calibrated cavities (0-45)

that fail to reflect green light

Count correct on a collection of 45theory calibrated test items

Substantive theory Thermodynamic theory Lexile Theory

Unit of measurement Degree Fahrenheit (oF) Lexile (L)

Correspondence table/calibration equation

Exploits a chemical reaction and lightabsorption to table temperature as a

function (Guttman Model)of a sufficient statistic

Exploits semantic and syntacticfeatures of test items to table reader

ability as a function (Rasch model) of a sufficient statistic

Measure/Quantity Measurement outcome converted intoa quantity via the substantive theory

Measurement outcome converted into a quantity via the substantive theory

Readable technology NexTemp Thermometer™ Oasis™

General objectivity Point estimates of temperature areindependent of the thermometer

Point estimates of reader ability areindependent of the reading test

Anatomy of Two Measurement Procedures

Page 10: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

10

Ten Features of Causal Response Models – whether Guttman or Rasch1. Both measurement procedures depend on within-person causal

interpretations of how these two instruments work. NexTemp uses a causal Guttman Model, The Lexile Framework for Reading uses a causal Rasch Model.

2. In both cases the measurement mechanism is well specified and can be manipulated to produce predictable changes in measurement outcomes (e.g. percent correct or percent of cavities turning black).

3. Item parameters are supplied by substantive theory and, thus, person parameter estimates are generated without reference to or use of any data on other persons or populations. Therefore, effects of the examinee population have been completely eliminated from consideration in the estimation of person parameters for reader ability and temperature.

Page 11: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

11

4. In both cases the quantitivity hypothesis can be experimentally tested by evaluating the trade-off property. A change in the person parameter can be off-set or traded-off for a compensating change in the measurement mechanism to hold constant the measurement outcome.

5. When uncertainty in item difficulties is too large to ignore, individual item difficulties may be a poor choice to use as calibration parameters in causal models. As an alternative we recommend, when feasible, averaging over individual item difficulties to produce “ensemble” means. These means can be excellent dependent variables for testing causal theories.

6. Index models are not causal because manipulation of neither the indicators nor the person parameter produces a predictable change in the measurement outcome.

Ten Features of Causal Response Models – whether Guttman or Rasch cont’d.

Page 12: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

12

7. Causal Rasch models are individual centered and are explanatory at both within-subject and between-subject levels. The attribute on which I differ from myself a decade ago is the same attribute on which I differ from my brother today.

8. When data fit a Rasch model differences between person measures are objective. When data fit a causal Rasch model absolute person measures are objective (i.e. independent of instrument).

9. The case against an individual causal account, although popular, has been poorly made. Investigators need only experiment to isolate the causal mechanism in their instruments, test for the trade-off property and confirm invariance over individuals. This has been accomplished for a construct, reader ability, that has been described by scholars as the most complex cognitive activity that humans regularly engage in. Given the success with reading, we think it likely that other behavioral constructs can be similarly measured.

10. Causal Rasch models make possible the construction of generally objective growth trajectories. Each trajectory can be completely separated from the instruments used in its construction and from the performance of any other persons whatsoever.

Ten Features of Causal Response Models – whether Guttman or Rasch, cont’d.

Page 13: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

13

To causally explain a phenomenon [a measurement outcome] is to provide information about the factors [person processes and instrument mechanisms] on which it depends and to exhibit how it depends on those factors. This is exactly what the provision of counterfactual information…accomplishes: we see what factors some explanandum M [measurement outcome, raw score] depends on (and how it depends on those factors) when we have identified one or more variables such that changes in these (when produced by interventions) are associated with changes in M (Woodward, 2003, p.204).

Page 14: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

14

How Many Ways Can We Say X Causes Y?

X “elicited a greater” Y

X “impacts” Y

X “accounts for” Y X “has been linked to” Y

Y “is the result of” X X “didn’t diminish” YY “because of” X Y “depends on” XX “has led to” Y X “largely motivates”

YY “stemmed from” X X “proved critical to”

YX “fosters” Y X “changes” YX “triggers” Y X “affects” Y

Page 15: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

15

Psychometrics vs. MetrologyAspect

Interpretation of ProbabilityGroup CenteredInterpretation involves 100

people with the same ability answering a single item

Individual CenteredInterpretation involves

administering 100 items with the same calibration to a single

person

Person Measures A person’s response record is embedded in different samples and each group specific Rasch analysis produces a different

measure

A person’s response record is evaluated against theory-

referenced calibrations

Measurement Error Traditional test theory uses a sample standard deviation and a sample correlation to compute

an SEM which is intended to characterize the individual

ISEM is the within person standard deviation over

replications of the measurement procedure

Data Fit to the Model Varies with the locally constructed frame of reference,

sample dependent

Fit is to a theory, thus, sample independent

Validity Correlational, thus, sample dependent

Causal within person, thus, sample independent

Page 16: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

16

Page 17: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

17

Page 18: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

18

r = 0.952

r” = 0.960

R2” = 0.921

RMSE” = 99.8L

Figure 1: Plot of Theoretical Text Complexity versus Empirical Text Complexity for 475 articles

“Pizza Problems”

Page 19: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

19

What could account for the 8% unexplained variance? Missing Variables Improved Proxies/Operationalizations Expanded Error Model Rounding Error Interaction between Individual and

Text Psychometric Uncertainty Principle

Page 20: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

20

Page 21: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

21

May 2016(12th

Grade)

Text Demands forCollege and Career

21

1200

1000

1400

1600

May 2007 – Dec. 2009284 Encounters117,484 Words2,894 Items848 Minutes

Student 15286th GradeMaleHispanicPaid Lunch

Page 22: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

22

Item-Based vs.Ensemble-Based Psychometrics

Page 23: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

23

Reading Task-Complexity Plane for Dichotomous Items

Native Lexile

Added Hardness

Added Easiness

Production Cloze

Auto-Generated

Cloze

1.31.2

1.1

1.0

0.9

0.8

0.7

Unit Size Adjustment Applied to Logits

Page 24: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

24

Comparing Item-Based vs. Ensemble-Based Psychometrics

Item-Based– Item statistics– Item characteristic curves– DIF for items

Ensemble-Based– Ensemble statistics– Ensemble characteristic curves– DIF for ensembles

Page 25: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

25

The Ensemble Objective: Correspondence Table

– Raw score to Lexile measure

What we think we know– Mean and spread of item distributions for a

passage

What is assumed to be unknown– Individual item difficulties

1300L(132L)

Page 26: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

26

The Process – Iteration 1

STEP 1

Sample 45 Item Difficulties from

Ensemble

STEP 2

Compute Lexile Measures for Each Raw Score (1 to 44)

STEP 3

Table Results

Raw Score

123...

44

Lexile Measure

362L514L584L

.

.

.1811L

Sample 1

Page 27: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

27

The Process – Iteration 2

STEP 1

Sample 45 Item Difficulties from

Ensemble

STEP 2

Compute Lexile Measures for Each Raw Score (1 to 44)

STEP 3

Table Results

Raw Score

123...

44

Lexile Measure

362L514L584L

.

.

.1811L

Lexile Measure

354L506L575L

.

.

.1797L

Sample 1 Sample 2

Page 28: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

28

The Process – Iteration 1,000

STEP 1

Sample 45 Item Difficulties from

Ensemble

STEP 2

Compute Lexile Measures for Each Raw Score (1 to 44)

STEP 3

Table Results

Raw Score

123...

44

Lexile Measure

362L514L584L

.

.

.1811L

Lexile Measure

354L506L575L

.

.

.1797L

Sample 1 … Sample 1,000

Mean Lexile Measure

378L509L589L

.

.

.1829L

Mean of 1,000

Page 29: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

29

ClosingNo matter how it is sliced and diced, analyses of joint and conditional probability distributions yield no more than patterns of association. Nothing in the response data nor Rasch analyses of these data exposes the processes (features of the object of measurement) or mechanisms (features of the instrument) that are hypothesized to be conjointly causal on the measurement outcomes.

Page 30: Jackson Stenner Chairman & CEO, MetaMetrics jstenner@lexile

30

A. Jackson Stenner CEO, [email protected]

Contact Info: