Windows Into The Mind Richard J. Shavelson & Maria Araceli - - PowerPoint PPT Presentation

windows into the mind
SMART_READER_LITE
LIVE PREVIEW

Windows Into The Mind Richard J. Shavelson & Maria Araceli - - PowerPoint PPT Presentation

Windows Into The Mind Richard J. Shavelson & Maria Araceli Ruiz-Primo Stanford University Invited Talk University of Ancona Ancona, Italy July ????, 2000 Assessing Knowledge Structure DENNIS THE MENACE GEE! IM GLAD WE DID CONCEPT


slide-1
SLIDE 1

Windows Into The Mind

Richard J. Shavelson & Maria Araceli Ruiz-Primo Stanford University

Invited Talk University of Ancona Ancona, Italy July ????, 2000

slide-2
SLIDE 2

July ????, 2000 Windows Into The Mind 2

Assessing Knowledge Structure

DENNIS THE MENACE “GEE! I’M GLAD WE DID CONCEPT MAPS - NOW WE BOTH KNOW WHAT I LEARNED TODAY!!! ”

slide-3
SLIDE 3

July ????, 2000 Windows Into The Mind 3

Overview

  • Cognitive theory underlying assessments of

knowledge structure

  • Direct and indirect methods of assessing

“cognitive structure”

  • Direct method: Concept maps
  • Indirect method: Cognitive maps
  • Link between concept and cognitive maps
  • Concluding comments
slide-4
SLIDE 4

July ????, 2000 Windows Into The Mind 4

Toward An Achievement Framework: Knowledge Types and Dimensions

Declarative Procedural Strategic Knowledge Knowledge Knowledge

(Knowing the “that”) (Knowing the “how”) (Knowing the “which,” “when,” and “why”)

Proficiency

Low High

Extent

(How much?)

Structure

(How is it organized?)

Others

(Precision? Efficiency? Automaticity?)

Cognitive Cognitive Tools: Tools:

Planning Planning Monitoring Monitoring

Domain-specific content:

  • facts
  • concepts
  • principles

Domain-specific production systems Problem schemata/ strategies/

  • peration systems
slide-5
SLIDE 5

July ????, 2000 Windows Into The Mind 5

Associative and Semantic Models of Memory

  • Nodes represent

concepts

  • Lines represent

relations among concepts

  • Probe memory by

asking about concepts (or lines)

Is a

Is an example of Contains

Associative Semantic

slide-6
SLIDE 6

July ????, 2000 Windows Into The Mind 6

What Are Concept Maps?

A Concept Map of a Concept Map!

graph

Concept Concept Map Map

consist of consist of is a

nodes labeled lines concepts relation

denote a represent linked by between a pair of

slide-7
SLIDE 7

July ????, 2000 Windows Into The Mind 7

Concept Map Assessment: An Example

The following map shows the relationship among concepts related to genetic continuity. The concepts are arranged hierarchically and linked each other. Please examine the map and supply, in the space provided, a word or two for labeling each link such that association between concepts is made clear.

Ayita, I think the student had to simply write in a number (1-5) on the line from words given

Task Task Response Format Response Format Scoring System Scoring System

Genetic continuity Asexual reproduction Sexual reproduction Meiosis Male gamete Female gamete

  • 1. through
  • 2. producer
  • 3. requires
  • 4. called
  • 5. fuse into
  • One point is given for each correct relationship
  • One point is given for each level up until two

levels beyond the last branching

  • One point is given for the first branching where

two or more concepts are connected to the concept

  • above. Three points are given for any subsequent

branching

  • One point is given for each cross link

Source: Tamir, P. (1995). Science Assessment. In M. Birenbaum & F. J. R. C. Dochy (Ed.). Alternatives in assessment of achievements, learning processes, and prior knowledge. Boston: Kluwer Academic Publishers.

slide-8
SLIDE 8

July ????, 2000 Windows Into The Mind 8

Are There Different Types of Concept Maps?

Yes, more than we would like! Yes, more than we would like! Variations in the task, the response format, and the scoring system produce different concept mapping techniques.

slide-9
SLIDE 9

July ????, 2000 Windows Into The Mind 9

Construct-A-Map

Construct a concept map that reflects what you know about what a concept map is. Organize the terms in relation to one another in any way you want.

concept map concepts relationships

composed of which elementary unit is composed of stand for with explanations show the essential relationship between composed of with explanations stand for

knowledge

can be represented to some extent by a determine the structure of a

linking lines nodes proposition

acquisition demands understanding the meaning of understanding demands clear explanation of

slide-10
SLIDE 10

July ????, 2000 Windows Into The Mind 10

Scoring System

Based on a criterion map, it focuses on:

  • Proposition validity:

Proposition validity: Accuracy of propositions

  • Congruence

Congruence: Proportion of valid propositions in student’s map to total possible valid propositions in criterion map

  • Salience:

Salience: Proportion of valid propositions to total number

  • f valid propositions in student’s map
  • Nodes:

Nodes: Concepts used in the student’s map

slide-11
SLIDE 11

July ????, 2000 Windows Into The Mind 11

Proposition “Quality”

Quality of Proposition Definition

Outstanding proposition. Complete and correct. It shows a deep understanding of the relation between the two concepts. acids-compounds: < that gives off H+ when dissolved in water are Excellent Excellent 4 points Good Good Complete and correct proposition. It shows a good understanding

  • f the relation between the two concepts.

acids-compounds: > are examples of

3 points Poor Poor

Incomplete but correct proposition. It shows partial understanding

  • f the relation between the two concepts.

acids-compounds: < form 2 points Don’t Care Don’t Care Although valid, the proposition does not show understanding between the two concepts. acids-compounds: > is a different concept 1 points Invalid Invalid Incorrect proposition. acids-compounds: > made of 0 points

slide-12
SLIDE 12

July ????, 2000 Windows Into The Mind 12

Fill-In-The-Map

Examine the map, the blank nodes and the terms provided on the list below. Select the term that corresponds to each node and write it down inside the circle. Use each term only once. List of Concepts:

  • concepts
  • graph
  • labeled lines
  • propositions

Score right or wrong

with

concept map

is a with

nodes

with represent linked by that describe the relationship between two nodes are called

slide-13
SLIDE 13

July ????, 2000 Windows Into The Mind 13

Fill-In-The-Linking-Lines Technique

Compounds Molecular Compounds Anions Ternary Ionic Compounds

ELEMENTS, ATOMS, AND COMPOUNDS

Atoms Elements are composed of the same type of Periodic Table are organized on a are negatively charged particles of have either a positive or a negative are formed when atoms lose or gain are combined chemically to form made up of

  • pposite

charged are composed

  • f more than
  • ne type of

is a ternary ionic compound that contains sulfite which is a are located

  • n the left

side of the are elements located in the center of the contain one

  • r more

with two elements are have to have hydrogen are composed of H+ and are composed of metals and tend to form are located on the upper right side

  • f the

have a positive have a negative will form Metals containing molecules and sharing electrons are called

C

are composed of two Transition Metals Non-Metals Charge Ions Polyatomic Ions Cations Electrons Ionic Compounds N2O4 Sodium Sulfite Binary Ionic Compounds Acids is an example of a with equal number of electrons and protons have a neutral have a negative

slide-14
SLIDE 14

July ????, 2000 Windows Into The Mind 14

On The Reliability & Validity of Concept Maps

  • Reliability: Can students’ concept maps be consistently scored by

different persons?

  • Equivalence of Mapping Techniques: Do all mapping techniques

provide the same picture about students’ declarative knowledge structure?

  • Task Demands: Do different mapping techniques impose different

cognitive demands to students?

  • Comparing Concept Maps With Traditional Assessments:

Do concept maps measure something different than multiple-choice test?

slide-15
SLIDE 15

July ????, 2000 Windows Into The Mind 15

Tasks

Mapping Technique 2 Mapping Technique 1

No-Concepts Provided Concepts Provided

  • Construct a concept map about Ions,

Molecules, and Compounds.

  • Select another 7 concepts that you

think are the most important in explaining the topic.

  • Organize the concepts
  • Construct a concept map about Ions,

Molecules, and Compounds.

  • Examine the 10 concepts and construct

the map.

  • Organize the concepts
  • Redraw the map so someone else can

read it.

slide-16
SLIDE 16

July ????, 2000 Windows Into The Mind 16

Generalizability(Reliability) of Scores

Percent of Total Variability and G Coefficients for Proposition Validity Total Scores No-Concepts, Sample A, and B Sample A & B Person (p) Rater (r) Sample (s) p x r p x s r x c prc,e ρ Φ ρ

2

Φ

2

71.64 78.67 .15 .79 22.81 17.64 .01 .18 5.37 2.69 .89 .88 .89 .88 .78 .78

(nr=1,ns=1) (nr=2,ns=2) (nr=2,ns=3)

slide-17
SLIDE 17

July ????, 2000 Windows Into The Mind 17

Discriminant Validity

Correlation Between Multiple-Choice Test Score and Concept Map Scores

Type of Score No-Concepts Sample A Sample B

Proposition Validity Congruence Salience .64 .66 .61 .63 .55 .50 .58 a .45

slide-18
SLIDE 18

July ????, 2000 Windows Into The Mind 18

Tasks

Construct-a-Map Fill-in-the-Linking Lines Fill-in-the-Nodes

  • Construct a map.
  • Use the 20 concepts

provided to construct the map.

  • Organize the conceptsany

way you want.

  • Check the map.
  • Redraw the map so

someone else can read it.

  • Examine the skeleton

map, the blank linking lines and the linking words provided.

  • Select the linking words

that correspond to each blank linking line.

  • Write the linking words
  • n the blank linking line.
  • Check that all blank lines

are filled-in.

  • Examine the skeleton

map, the blank nodes and the terms provided.

  • Select the term that

corresponds to each blank node.

  • Write the term inside the

circle.

  • Check that all blank

nodes are filled-in.

slide-19
SLIDE 19

July ????, 2000 Windows Into The Mind 19

Response Format

Construct-a-Map Fill-in-the-Linking Lines Fill-in-the-Nodes

Blank page to construct the map

Elements, Atoms, and Compounds

atoms compounds N2O4

made up

  • f opposite

composed of with 3 or more elements are

ions ionic compounds molecular compounds ternary ionic

acids formed from molecules sharing

atoms compounds N2O4

Elements, Atoms, and Compounds

formed with cations and anions are formed from molecules sharing example of made up

  • f opposite

composed of that begins with H are

lose or gain electrons to form

with 3 or more elements are

slide-20
SLIDE 20

July ????, 2000 Windows Into The Mind 20

Are Fill-In-The-Map Scores Sensitive To The Nodes/Linking Lines Selected To Be Filled-In?

Means and Standard Deviations by Type of Map and Sample

Mean S.D.

Type of Skeleton Map Fill-in-the-nodes Sample 1 80 11.21 1.42 Sample 2 72 10.80 1.74 Fill-in-the-linking lines Sample 1 78 9.77 2.74 Sample 2 73 8.99 3.09

(Max = 12)

n

slide-21
SLIDE 21

July ????, 2000 Windows Into The Mind 21

Are The Two Forms Of Fill-In- The-Map Equivalent?

Means and Standard Deviations by Type of Map and Sequence

Fill-in-the-linking lines Fill-in-the-nodes Sequence

Mean S.D. Mean S.D. n

1 Nodes 1-Lines 1 43 11.09 1.52 9.72 2.84 2 Nodes 1-Lines 2 36 11.03 1.33 9.31 3.06 3 Nodes 2-Lines 1 35 10.63 1.81 9.83 2.65 4 Nodes 2-Lines 2 37 10.97 1.67 8.68 3.13 Total 152 11.02 1.59 9.36 2.93

slide-22
SLIDE 22

July ????, 2000 Windows Into The Mind 22

Can The Two Mapping Techniques Be Considered Equivalent?

Means and Standard Deviations across the Three Types of Assessments

Type of Assessment

Max n Mean S.D.

Construct-A-Map

Proposition Accuracy 152 135 53.91 22.17 Convergence 152 1 .50 .19 Salience 152 1 .73 .17

Fill-In

Fill-in-the-nodes 152 12 11.02 1.59 Fill-in-the-lines 151 12 9.39 2.93

Multiple-Choice Test

150 30 24.05 3.74

slide-23
SLIDE 23

July ????, 2000 Windows Into The Mind 23

Do The Mapping Techniques Provide Similar Information Than The One Provided By A Multiple-Choice Test?

Type of Assessment

Construct-A-Map (C-M)

(.99)

Fill-in-the-nodes-NOD (FI-N)

.47 (.70)

Fill-in-the-lines-LIN (FI-L)

.44 .40 (??)

Multiple-Choice Test (MC)

.44 .37 .53 (.74) Structure Amount C-M FI-N FI-L MC

slide-24
SLIDE 24

July ????, 2000 Windows Into The Mind 24

Some Empirical Evidence: Other Evidence

  • Correlations (covariances) tell you about relative

rank ordering on different measures--that’s only part of the story

  • Evidence is needed for the cognitive claims that

different methods measure somewhat different aspects of achievement:

– Talk aloud – Focus group – Group work – Other?

slide-25
SLIDE 25

July ????, 2000 Windows Into The Mind 25

Talk Aloud with Concept Maps: Method Variation

0.1 0.2 0.3 0.4 0.5 0.6

Explanation Monitoring Conceptual Errors No-Code

Contruct-A-Map Fill-In-Nodes Fill-In-Lines

Verbal Codes Proportion Scores

Correspondence Between Directedness and Inferred Activity

Low Directed High Directed

slide-26
SLIDE 26

July ????, 2000 Windows Into The Mind 26

Tentative Concept Maps Conclusions

  • Students’ maps can be consistently scored by different persons even

when complex judgments, such as proposition quality, are required

  • Different mapping methods provide different pictures of students’

declarative knowledge structure. The construct-a-map method provides

  • pportunities to reveal students’ conceptual understanding
  • Different mapping techniques impose different cognitive demands on
  • students. Highly structured mapping techniques allow students to

respond by elimination or guessing, whereas constructed response do not.

  • Correlations between concept-map and multiple-choice scores are

positive and moderate suggesting that these two types of assessment measure overlapping but somewhat different aspects of declarative knowledge

slide-27
SLIDE 27

July ????, 2000 Windows Into The Mind 27

Cognitive Maps

  • Methods

– Word Association – Similarity Judgment – Card Sorting – Tree Building

  • Data structure: Proximities or distances
  • Scaling methods: Networks or hierarchical

clusters

slide-28
SLIDE 28

July ????, 2000 Windows Into The Mind 28

Example: Similarity Judgments

Rate the similarity of each pair of terms by circling

  • ne of the numbers provided.
  • 1. Central Tendency

Mean

(Closely Related)

1 2 3 4 5 6 7

(Unrelated)

DK

(Circle one number -or- "DK" (Don't Know))

  • 2. Hypothesis

Description

(Closely Related)

1 2 3 4 5 6 7

(Unrelated)

DK

(Circle one number -or- "DK" (Don't Know))

slide-29
SLIDE 29

July ????, 2000 Windows Into The Mind 29

Small Pilot Study (N=28) Of Similarity Judgments And:

  • Concept maps
  • Multiple choice

Terms: Variability, Variance, Standard Deviation Variability Variance

  • Std. Dev.

measures measures is the square root of

  • 1. Which of the following refers

to a procedure for making generalizations from a limited random sample of data?

  • a. Statistical Inference
  • b. Intuition
  • c. Population Parameterization
  • d. Standardization
slide-30
SLIDE 30

July ????, 2000 Windows Into The Mind 30

Sample Portion Of Concept Map

uses sample to test ideas in the is used to test ideas about population in is used to test hypotheses in is used to test population-related addresses ideas about variables in population in the process of i s a p r

  • c

e d u r e i n s t a t i s t i c a l is one special case of the general linear model, along with i s a p r

  • c

e d u r e i n s t a t i s t i c a l refers to relationships within the

Hypothesis Sample Linear Regression Population Inference ANOVA

slide-31
SLIDE 31

July ????, 2000 Windows Into The Mind 31

Proximity Matrix For A Student

slide-32
SLIDE 32

July ????, 2000 Windows Into The Mind 32

Network Representation of Proximities

slide-33
SLIDE 33

July ????, 2000 Windows Into The Mind 33

Correlation Among Methods

MC CM Pretest CM 0.155 SJ 0.554 ** 0.251 ** p<.01 MC CM Posttest CM 0.330 SJ 0.434 * 0.706 ** ** p<.01

slide-34
SLIDE 34

July ????, 2000 Windows Into The Mind 34

What If We Could?

  • Collect cognitive structure information

indirectly by computer

  • Immediately show a student a

representation of her knowledge structure

  • Permit her to modify the structure
  • And label the lines?
slide-35
SLIDE 35

July ????, 2000 Windows Into The Mind 35

Conclusions

  • Concept and cognitive maps appear to tap a

different aspect of declarative knowledge than do multiple-choice tests

  • Concept and cognitive maps may provide

similar representations of structure although this has not been tested directly

  • By assessing different aspects of knowledge,

we may broaden what gets taught!