Testing - AMT - Chapter 9 Flashcards

1
Q

According to the authors, the best way to test people’s writing ability is….

A

to get them to write (test writing ability directly)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Writing task problems

A
  • Representative Tasks (population + expected performance)
  • Valid samples of writing (represent ability)
  • Scored Validly (Can/Will be)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

How to Create Representative Writing Task

A
  • Specify all possible content
  • Include Representative Sample of Specified Content
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Specify all possible content of a Writing Task

A

Operations (Describe, explain, compare, etc.)

Types of text

Addressees

Length of texts

Topics

Dialect and style

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Text Operations

(e.g.)

A

Describe, explain, compare and contrast, argue for and against a position, etc.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

ideal writing test for Construct Validity

A

One which requires candidates to perform all the relevant potential writing tasks.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Why writing task with

  • Wide ranging
  • Representative sample of specifications
A

++Validity

++Backwash Effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Really good coverage of the range of potential writing tasks is…

A

Not possible in a single version of the test.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

To know something ACCURATE and MEANINGFUL about a person’s WRITING ability is ALWAYS

A
  • Expensive
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Elicit a valid sample of writing ability

A
  • Separate tasks (as many as is feasible)
  • Test ONLY writing ability
  • Restrict candidates
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Ensure valid and reliable scoring

A
  • Reliably scored tasks
  • As many as possible tasks
  • Restrict candidates (directly comparable)
  • Long enough Samples
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Scales of Scoring (approaches)

A

Holistic & Analytic

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Scales of Scoring

(where in specifications)

A

Criterial Levels of Performance

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

”"”impressionistic”” refers to”

A

Holistic scoring

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Holistic Scoring

A
  • Single Score
  • Overall Impression
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Holistic Scoring

  • Advantages
A

Very Rapid (2min)

17
Q

Holistic Scoring

  • disadvantages
A
  • Not reliable
  • if scored 4 times = reliability improved
18
Q

Analytic Scoring

A

Seperate Score for each Aspects of a Task

19
Q

Analytic Scoring

  • Advantages
A
  • Subskills development in individuals considered
  • All Aspects of performance considered by Scorer
  • Number of Scores = ++Reliable
  • Weight of grade adapted
20
Q

Analytic Scoring

  • Disadvantages
A
  • Time consumming to score
  • Attention to different aspects distract from overall effect
  • Frequency of errors vs. Communication
21
Q

If diagnostic information is required directly from the ratings given

(Analytic or Holistic??)

A

Analytic scoring is essential

22
Q

Calibrate the scale to be used

A

1 - Collecting samples of performance

2 - Expert assing sample to point(s) on relevant scale

3 - Result = reference points

23
Q

Training Stages

A

1 - Background and Overview (study handbook)

2 - Rate + Discuss (handbook answered)

3 - Agreement on how to rate