HP3804 NTU Psychology Statistics: Module Review (Introduction to Psychometrics)

Note: this post is part of a series of posts regarding HP3804 (Psychological Testing)

What is Psychometrics (Psychological Testing)? Psychometrics — at its core — is all about measurement. It is the subfield in Psychology devoted to developing reliable tools to measure constructs like extroversion, risk-propensity, creativity, intelligence etc.

Unlike in the physical sciences where tools are rather clear-cut (ruler for length, weighing scale for weight), the field of Psychology is troubled by the fact that our constructs and their measurement can seem a bit airy/subjective. Take for instance extroversion — what is extroversion? How do you objectively measure extroversion? How do you come up with a score — whereby a higher score means that an individual is more extraverted than another?

Typically — we measures our constructs with scales (questionaires) in Psychology. But this raises new concerns too — what is the basis of this scale? What makes this scale better than just a layman’s subjective judgement of “oh I think this guy is extroverted”?

Psychological Testing in Everyday Life

If you think about it — the concept of measuring “un-concrete contructs” is not foreign to you. Especially in Singapore (where we love to take test papers and chase paper qualifications) — examinations are a form of Psychometric Test! Things like PSLE, O Levels, A levels — they are basically Psychological Tests attmpting to quantify “mastery of the course syllabus” if you willl.

The more general form of these assessments would be intelligence assessments— like IQ tests. Intelligence, as a construct, also falls sqaurely under the domain of Psychology (as it is not clear cut as to what it actually is lol). There are many domains of intelligence (many models of intelligence in fact) — which differ in terms of how they understand the construct (e.g. is it a singular construct — Spearman’s General Intelligence? Are there fundamentally different forms of intelligence — like analytical, creative, practical intelligence — in which indiviuals can vary upon independently from each other — Sternberg Tricarhcic Theory?)

Even after settling the fundamental understanding of the construct — we still face yet another problem. How do we measure these constructs? What questions do we set that can accurately capture a person’s proficiency in these domains? How many questions do we need? Is a single questoin enough? Can we account for careless mistakes in our assessment?

Side note: we need to care about these things because the goal is to accurately measure a person — we want our assessment to be robust enough to still give a smart person a high score — even if he/she sometimes makes mistakes.

History of Psychological Testing

Believe it or not, one of the earliest known forms of psychological testing originated in China. Historically, to become an official in the imperial court, an individual first had to pass the Civil Service Examination. The goal of this examination was to select candidates who were intelligent, capable, and well-educated — essentially, those most likely to succeed in government roles.

The Civil Service Exam tested not only knowledge of Confucian classics and literature but also the candidate’s reasoning, memory, and analytical skills. In many ways, this system can be seen as a precursor to modern psychological testing because it attempted to objectively measure abilities and aptitudes to predict future performance in a professional role.

Overt time, this idea extended to other fields like the military (selecting capable fighter pilots who perofmr well under pressure), doctors (selecting ethical and capable doctors), police (selecting individuals with a strong sense of justice and morality) — and now almost every job (accountant, HR, service staff) — all undergo some form of assessment for relevant skills before employment.

Psychological Tests as the Backbone of Psychology

Psychometrics is upstream of the study of Psychology. It is a requirement — for the study of Psychology is research based. In Psychology, we try to prove relationships between constructs — whether extroversion is linked to intellignence, salary is linked to well being, emotional intlligence linked to success in the workplace etc.

Before you can even talk about relationships between variables — you have to be able to capture the variable in the first place. Which is why the very first thing you learn in Psychology (HP1100 Fundamentals of Social Science Research!) is about reliability and validity (albeit very briefly).

2 Types of Tests: Ability vs Attitude Tests

In Psychometrics — we generally distinguish between 2 types of tests

1. Ability tests — where questions have a right anwer
2. Personality/Attitude tests — where questions where do not have a right answer

The framework for development and quantifying these scores are very different (naturally). The former is more commonly seen in educatoinal settings (exam paper setting) — whereas the latter is more broadly applicable across varies subdomains of Psyhcology.

In HP3804 — we are primarily focusing on the 2nd type of scales — where questions have no right answer (e.g. How much do you agree the statement: “I enjoy constantly being around people?”).

Question Formats

There are numerous ways one can ask the same question inside a questionaire. Some options include:

1. True/False Question (I enjoy public speaking)
2. Ranking Questions (Rank the activities from most to least preferred: Public Speaking, Taking Examinations, Sleeping)
3. Likert Scale Type Questions (From a scale of 1–10, how much do you enjoy public speaking?)
4. Negative Questions (How much do you hate public speaking?)
5. Short answer question (Elaboate on how you feel about public speaking)

For the most part, we use Likert Scale Type Questions. And indeed, majority of the things we learn in HP3804 concern analysis of these Likert Scale Type Questions (each other question format actually requires their own specialised from of analysis method).

Conclusion

The stage has been set — we now have a clearer picture of what Psychometrics tries to achieve. From ancient examinations to modern psychological tests, the theme has always been the same: how do we measure something invisible?

With that in mind, it’s time to revisit a deceptively simple question that underpins all of Psychology — one that we glossed over back in our first-year modules: what is a construct? The answer might sound straightforward, but it’s the cornerstone of every test, every questionnaire, and every analysis that follows. Stay tuned!