0% found this document useful (0 votes)
6 views30 pages

Notes Testing

The document outlines the process of test construction, including its importance, steps, and various scaling methods used to measure psychological constructs. It also covers case history taking, emphasizing the skills needed for effective client interviews and the information to collect. Additionally, it discusses reliability and validity theories, detailing how to assess test consistency and accuracy, and the significance of norms in interpreting test scores.

Uploaded by

Haniya Sadiq
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
6 views30 pages

Notes Testing

The document outlines the process of test construction, including its importance, steps, and various scaling methods used to measure psychological constructs. It also covers case history taking, emphasizing the skills needed for effective client interviews and the information to collect. Additionally, it discusses reliability and validity theories, detailing how to assess test consistency and accuracy, and the significance of norms in interpreting test scores.

Uploaded by

Haniya Sadiq
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

🧠 1.

TEST CONSTRUCTION (MOST IMPORTANT)

🔹 What is Test Construction?

It’s the process of creating a psychological test to measure something (e.g., intelligence,
anxiety).

👉 Why create a new test?

 No test exists

 Existing tests are not good enough

🔹 6 STEPS OF TEST CONSTRUCTION (VERY IMPORTANT – MEMORIZE)

1. Defining the Test

👉 Decide:

 What are you measuring?

 Why?

 How is it different from existing tests?

📌 Example:
If measuring intelligence:

 Focus on problem-solving vs memory

 Make it useful for education

2. Selecting a Scaling Method

👉 How will you assign numbers to responses?

🔸 LEVELS OF MEASUREMENT (SUPER IMPORTANT QUESTION AREA)

1. Nominal Scale

 Just labels/categories

 No order
📌 Example:

 Male = 1, Female = 2
❗ Numbers have NO meaning

2. Ordinal Scale

 Ranking/order

 No equal gaps

📌 Example:

 1st, 2nd, 3rd position


❗ You don’t know how much difference

3. Interval Scale

 Equal differences between numbers

 No true zero

📌 Example:

 Temperature

 Rating scale (1–100)

4. Ratio Scale

 Has true zero

 Can say “twice as much”

📌 Example:

 Weight

 Number of siblings

🔥 Exam Tip:
Scale Key Feature

Nominal Labels only

Ordinal Rank

Interval Equal gap

Ratio True zero

🔸 Scaling Methods

1. Expert Ranking

 Experts list behaviors

 Rank them from mild → severe

📌 Example: Depression scale

2. Likert Scale (VERY COMMON)

 Strongly Agree → Strongly Disagree

 Scores: 1–5

📌 Total score = sum of all items

3. Empirical Keying

 Based on data, not theory

👉 Compare:

 Anxiety group vs Normal group

👉 Keep items that:

 Only anxious people agree with

🔹 3. Constructing Items

👉 Writing questions
Questions to consider:

 Easy or difficult?

 How many?

 What format?

🔹 4. Testing the Items (VERY IMPORTANT)

This is ITEM ANALYSIS

Key Terms:

✔ Item Difficulty Index

 How easy/hard a question is

👉 Good test:

 Mix of easy + hard

✔ Ceiling Effect

 Everyone scores very high

✔ Floor Effect

 Everyone scores very low

✔ Item Discrimination Index

 Can the question distinguish:

o Smart vs weak students?

👉 Good item:

 High scorers get it right

 Low scorers get it wrong

✔ Item Characteristic Curve (ICC)


 Graph of:

o Ability vs probability of correct answer

🔹 5. Revising the Test

 Remove bad questions

 Improve test

 Test again on new sample

🔹 6. Publishing the Test

 Create manuals

 Make materials user-friendly

🧠 2. CASE HISTORY TAKING

👉 This is about interviewing a client

🔹 Important Skills

1. Body Language

 Lean forward = interest

 Don’t cross arms ❌

 Nod and smile (not too much)

2. Verbal Cues

 Speak clearly

 Avoid:

o “umm”

o “like”
o “you know”

🔹 INFORMATION TO COLLECT (VERY IMPORTANT)

✔ Identifying Information

 Name, age, gender

 Education

 family, siblings

 SES (socioeconomic status)

✔ Presenting Complaint

 Main problem

📌 Example:
“Feeling anxious for 2 months”

✔ General Health

 Physical + mental illness

 Medication

✔ Academic History

 Grades

 Behavior

 Friends

✔ Work History

 Jobs

 Performance
 Issues

✔ Family History

 Relationships

 Home environment

✔ Social History

 Friends

 Activities

✔ Personality

 Strengths & weaknesses

✔ Other Important Areas

 Mood

 Sleep

 Memory

 Drug use

🔥 KEY LINE FOR EXAM:

👉 Always observe client’s behavior during sessions

🧠 3. RELIABILITY THEORIES (VERY IMPORTANT THEORY QUESTION)

👉 Reliability = consistency of test scores

🔹 Classical Test Theory (CTT)


Formula:

👉 Observed Score = True Score + Error

👉X=T+E

 Error = unknown

 Problem: Doesn’t explain error

🔹 Domain Sampling Theory

👉 Test = sample of all possible questions

📌 Example:

 50 questions represent a huge syllabus

👉 Reliability depends on:

 How well sample represents whole

📌 Analogy:
Soup tasting 🍲

🔹 Generalizability Theory (G-Theory)

👉 Breaks error into parts

Instead of:

 “There is error”

It says:

 Error from examiner?

 Environment?

 Test?

📌 Very detailed analysis

🔹 Item Response Theory (IRT) (VERY IMPORTANT)


👉 Focuses on individual questions

Considers:

 Difficulty

 Discrimination

 Guessing

📌 Used in:

 GRE

 Adaptive testing

👉 Questions change based on your answers

🔥 FINAL COMPARISON (MEMORIZE THIS)

Theory Idea

CTT Score = true + error

Domain Sampling Test = sample

G-Theory Finds source of error

IRT Focus on each item

🚨 LAST MINUTE REVISION (READ THIS BEFORE EXAM)

MOST IMPORTANT TOPICS:

✅ Steps of test construction


✅ Levels of measurement
✅ Likert scale
✅ Item difficulty + discrimination
✅ Case history components
✅ CTT formula
✅ IRT vs CTT

⚡ QUICK MEMORY HACKS


 Nominal = Name

 Ordinal = Order

 Interval = Equal gap

 Ratio = Real zero

 CTT = Basic formula

 IRT = Smart questions

BIG PICTURE (VERY IMPORTANT)

All three topics are connected:

 Reliability → “Is the test consistent?”

 Validity → “Is the test measuring the right thing?”

 Norms → “How do we interpret the score?”

👉 Think:

Reliable ≠ Valid
But Valid test MUST be reliable

🧠 PART 1: RELIABILITY (Consistency)

✅ What is Reliability?

 It means trustworthiness of test scores

 If you repeat the test → results should be similar

👉 Example:
You take IQ test today = 110
Next week = 112
✔ Reliable (close)

But if:
110 → 80
❌ Not reliable
🔑 TWO CORE IDEAS

1. Consistency → same results again

2. Precision → scores are accurate (not random)

📊 Reliability Coefficient (rₓₓ)

 Range: 0 → 1

 Closer to 1 = high reliability

Value Meaning

0 No reliability

0.5 Moderate

0.9+ Very high

📈 Correlation (VERY IMPORTANT)

Reliability is measured using correlation

Types:

 Positive (+) → both increase

 Negative (–) → one up, one down

 Zero → no relation

🔧 TYPES OF RELIABILITY

1. Inter-Rater

 Different examiners give similar scores


👉 Example: Essay grading

2. Test-Retest

 Same test → two times


 Measures stability over time

⚠️Issues:

 Too short gap → memory effect

 Too long gap → real changes

3. Alternate Forms

 Two versions of same test

 Avoids memorization

4. Internal Consistency

 Are all questions measuring same thing?

✔ Measured by:

 Cronbach’s Alpha

 K-R 20

5. Split-Half

 Divide test into two halves

 Compare both

💡 KEY EXAM LINE

Reliability = consistency of scores across time, items, or raters

🎯 PART 2: VALIDITY (Accuracy)

✅ What is Validity?

Does the test measure what it CLAIMS to measure?


👉 Example:
Math test that includes history questions
❌ Not valid

🔥 IMPORTANT POINTS

 Validity is not yes/no → it’s a degree

 A test is valid only:

o for a specific purpose

o for a specific group

⚠️GOLDEN RULE

A test can be reliable but NOT valid


But a valid test MUST be reliable

🧪 TYPES OF VALIDITY

1. Content Validity

👉 “Does test cover ALL parts of the topic?”

Example:

 Math test should include:

o algebra

o geometry

o arithmetic

✔ If complete → high content validity

How measured?

 Expert judges review items


2. Criterion-Related Validity

👉 Compare test with real-world outcome (criterion)

(i) Concurrent Validity

 Measured at same time

Example:
New depression test vs existing test

(ii) Predictive Validity

 Predicts future performance

Example:

 Entry test → predicts job success

3. Construct Validity

👉 Does it measure an abstract concept?

Examples:

 Intelligence

 Anxiety

 Self-esteem

(i) Convergent Validity

 Should correlate with similar constructs

👉 Example:
IQ test ↔ another IQ test

(ii) Discriminant Validity


 Should NOT correlate with unrelated things

👉 Example:
IQ test ≠ height

😎 Bonus: Face Validity

 “Looks valid”

 Not scientific, just perception

💡 KEY EXAM LINE

Validity = how well a test measures what it is intended to measure

📊 PART 3: NORMS (Meaning of Scores)

❗ KEY IDEA

Raw scores are USELESS alone

👉 Example:
You got 70 marks
So what?
Is it good or bad?

👉 You need comparison

👥 Norm Group

 A representative sample

 Used to compare scores

📈 Norms

 Scores based on group performance

👉 Tell you:
 low

 average

 high

📊 DESCRIPTIVE STATISTICS (IMPORTANT)

1. Frequency Distribution

 Table showing how many scores fall in ranges

2. Histogram

 Graph version of frequency distribution

3. Mean

 Average score

4. Standard Deviation (SD)

👉 Spread of scores

 Small SD → scores close together

 Large SD → spread out

🔔 Normal Distribution

 Bell-shaped curve

 Most scores in middle

🔄 Skewness
Type Meaning

Positive skew more low scores

Negative skew more high scores

📊 TRANSFORMED SCORES

1. Percentile Rank

👉 Position, NOT marks

Example:

 90th percentile = better than 90% students

2. Standard Score (z-score)

👉 Distance from mean

 Positive → above average

 Negative → below average

🧪 TYPES OF NORMS

1. Developmental Norms

Age Norms

 Compare same age

Grade Norms

 Compare same class

2. Within-Group Norms

Local Norms
 Specific area (e.g., Karachi)

Subgroup Norms

 Specific group (e.g., females)

🎯 Criterion-Referenced Tests

👉 Compare with fixed standard, not people

Example:

 IELTS

 GRE

✔ Focus: “Can you do it?”

⚡ LAST-MINUTE REVISION (SUPER IMPORTANT)

🔁 Reliability

 Consistency

 Measured by correlation

 Types: test-retest, inter-rater, etc.

🎯 Validity

 Accuracy

 Types:

o Content

o Criterion (predictive + concurrent)

o Construct (convergent + discriminant)

📊 Norms

 Give meaning to scores


 Include percentiles, z-scores

 Based on norm group

Got you — let’s slow this down and make everything crystal clear with simple explanations +
real examples so you can actually understand (not just memorize).

🧠 PART 1: CASE HISTORY DOMAINS (DETAILED + EXAMPLES)

Think of case history as:


👉 “Collecting the full life story of a client”

🔹 1. Identifying Information

Basic facts about the person.

📌 Example:

 Name: Ali Khan

 Age: 21

 Gender: Male

 Education: BS Psychology

 Family: 2 siblings

 SES: Middle class

👉 WHY important?

 Helps understand background + context

🔹 2. Presenting Complaint (VERY IMPORTANT)

👉 The main problem the client comes with

📌 Example:

 “I feel anxious all the time”

 “I cannot sleep at night”

 “I get panic attacks before exams”


👉 Write in:

 Client’s own words (best practice)

🔹 3. General Health Condition

👉 Physical + mental health

📌 Example:

 Has asthma

 Taking antidepressants

 History of migraines

👉 WHY important?

 Physical issues can affect mental state

🔹 4. Academic History

👉 School/college life

📌 Example:

 Good grades till grade 8, then decline

 Bullied in school

 No participation in activities

👉 What to look for:

 Stress, trauma, performance issues

🔹 5. Work History

👉 Jobs + workplace behavior

📌 Example:

 Worked 2 years as teacher

 Left job due to stress


 Conflict with boss

👉 Helps identify:

 Responsibility level

 Stress tolerance

🔹 6. Family History (VERY IMPORTANT)

👉 Relationships at home

📌 Example:

 Strict father

 Supportive mother

 Frequent fights at home

👉 WHY important?

 Many psychological issues come from family

🔹 7. Social History

👉 Friendships + social life

📌 Example:

 Few friends

 Avoids gatherings

 Feels lonely

🔹 8. Personality (Strengths & Weaknesses)

📌 Example:

 Strength: hardworking

 Weakness: overthinking
🔹 9. Mood, Sleep, Memory, etc.

📌 Example:

 Mood: sad most days

 Sleep: 4 hours only

 Appetite: low

 Memory: forgetful

🔥 KEY TIP:

👉 Always observe behavior:

 Eye contact

 Nervousness

 Confidence

🧠 PART 2: TEST CONSTRUCTION (WITH EASY EXAMPLES)

🔹 STEP 1: Defining the Test

👉 Example:
You want to make an anxiety test

You decide:

 Measure exam anxiety

 For university students

🔹 STEP 2: Scaling Methods (WITH EXAMPLES)

✔ Nominal Scale

👉 Just categories
📌 Example:
“Have you experienced anxiety?”

 Yes = 1

 No = 2

❗ No ranking

✔ Ordinal Scale

👉 Ranking

📌 Example:
Rank stress level:
1 = Low
2 = Medium
3 = High

❗ We don’t know exact difference

✔ Interval Scale

👉 Equal gaps

📌 Example:
Rate anxiety:
0–100 scale

👉 Difference between 20–30 = same as 70–80

✔ Ratio Scale

👉 True zero

📌 Example:
“How many panic attacks per week?”
0 = none

🔹 Scaling Methods (IMPORTANT)


✔ Likert Scale (MOST COMMON)

📌 Example question:
“I feel nervous before exams”

Options:
1 = Strongly disagree
2 = Disagree
3 = Neutral
4 = Agree
5 = Strongly agree

👉 If student chooses 5 → high anxiety

👉 Total score = sum of all answers

✔ Expert Ranking

📌 Example:
Experts list anxiety symptoms:

 Sweating

 Heart racing

 Avoidance

Then rank:
Mild → Severe

✔ Empirical Keying (EASIEST WAY TO UNDERSTAND)

👉 Compare 2 groups:

Group A: Anxious people


Group B: Normal people

📌 Question:
“I avoid social situations”

 Group A: says YES


 Group B: says NO

👉 This question is GOOD → keep it

🔹 STEP 3: Constructing Items

👉 Example:
Write many questions like:

 “I feel nervous in crowds”

 “I overthink small things”

👉 Make:

 Easy + hard questions

🔹 STEP 4: Item Analysis (VERY IMPORTANT)

✔ Item Difficulty

📌 Example:
If 90% students answer correctly → TOO EASY
If 10% → TOO HARD

👉 Best = moderate difficulty

✔ Item Discrimination

👉 Does question separate smart vs weak?

📌 Example:

 Top students → correct

 Low students → wrong

👉 Good question ✔

✔ Ceiling Effect
📌 Example:
Everyone scores 95%
👉 Test too easy ❌

✔ Floor Effect

📌 Example:
Everyone fails
👉 Test too hard ❌

🔹 STEP 5: Revising

👉 Remove bad questions

🔹 STEP 6: Publishing

👉 Final test + manual

🧠 PART 3: RELIABILITY THEORIES (SIMPLE + CLEAR)

This is where you’re confused — so let’s make it VERY EASY.

🔹 What is Reliability?

👉 Consistency of scores

📌 Example:
If you take test today & tomorrow → same result = reliable

🔴 1. CLASSICAL TEST THEORY (CTT)

👉 Formula:
👉 Observed Score = True Score + Error

Think like this:


📌 Example:
You score = 80

But:

 True ability = 85

 Error = -5

❗ Problem:

👉 We don’t know:

 Where error comes from

🟡 2. DOMAIN SAMPLING THEORY

👉 Key Idea:
Test = small sample of huge content

📌 Example:
Exam of 50 questions
But syllabus has 500 topics

👉 So:
Test ≠ full knowledge
Only a sample

🔥 Important:

👉 Reliability depends on:

 How well test represents whole syllabus

📌 Easy Analogy:
🍲 Soup tasting
One spoon = whole pot
🟢 3. GENERALIZABILITY THEORY (G-THEORY)

👉 This improves CTT

Instead of:
“Error exists”

👉 It says:
“Let’s find WHERE error comes from”

📌 Example:
Student fails exam — why?

 Teacher strict?

 Bad environment?

 Poor questions?

👉 G-theory separates these errors

🔥 Key Word:

👉 “Facets” = sources of error

🔵 4. ITEM RESPONSE THEORY (IRT) (VERY IMPORTANT)

👉 Focus = individual questions

Key Idea:

👉 Not all questions are equal

📌 Example:
Question 1 = very easy
Question 2 = very hard

IRT says:
👉 Score should depend on difficulty
3 MAIN FACTORS:

1. Difficulty

Hard or easy?

2. Discrimination

Does it separate smart vs weak?

3. Guessing

Can people guess correctly?

📌 Example:
MCQ → guessing possible

🔥 BEST EXAMPLE:

Computer-based test

👉 If you answer correctly:

 Next question becomes harder

👉 If wrong:

 Next becomes easier

🔥 FINAL SUPER SIMPLE SUMMARY

Theory Simple Meaning

CTT Score = true + unknown error

Domain Test = sample

G-Theory Breaks error into parts

IRT Focus on each question


⚡ WHAT TO DO NOW (IMPORTANT)

👉 Spend next 30–40 mins:

1. Read summaries again

2. Memorize tables

3. Practice examples

If you want, I can:


✅ Give predicted exam questions
✅ Do a rapid-fire quiz (best for memory)
✅ Make a 1-page cheat sheet

Just tell me 👍

You might also like