Most studies use the NEO-PI-R- 240 item scale by Costa & McCrae (1992) OR NEO-FFI- 60 item scale by Costa & McCrae (2003). Despite being the literature has many other shorter and new scales. The newest and shortest one is the Ten Item Personality Inventory (TIPI) by Gosling, Rentflow & Swann (2003). I am looking for an explanation on why using the NEO scale and why not using other scales like TIPI.
Some scales are more carefully constructed than others. Costa & McCrae are true experts in the five-factor model and good professional test developers. To be fair, if all one wants is a basic five-factor profile the complete NEO-PI-R (or newer NEO-PI-III) is unnecessary; the NEO-FFI is perfectly adequate to one's needs. (The longer test includes 5-6 subscales for each major trait, so it is preferable when a real in-depth personality profile is needed.)
The ten-item test? Well, you have just 2 questions per scale, obviously. This despite the fact that the "Big Five" traits are complex, wide-ranging constructs. Can 2 items capture the whole of Extraversion? Or Neuroticism. Short answer: no. The most one can hope for is 2 items that either cover different aspects of the trait (but don't correlate very well with each other) or correlate well with each other (because they are limited to one facet of the trait).
For the first scenario, I might write (1) I like to meet a lot of new people and (2) My opinion usually carries a lot of weight in group meetings. For the second, I might write (1) I like to meet a lot of new people and (2) I have a wide circle of acquaintance.
Bottom line: you lose breadth and depth of coverage. You lose reliability. But if the items are good ones (and the TIPI isn't bad, as far as it goes) you still get halfway decent data.
Why do people use a personality trait measure with just 2 items per scale? Because they are running complex multivariate analyses of data sets with many cases, so they want to skimp as much as possible on each measure - and you can't have a variable in (say) a structural equation model unless you have two (2) measures per variable. Otherwise they'd settle for one item.
So if you're running a big study (hundreds upon hundreds of participants) with lots of measures per person, and the Big Five traits aren't the "main event" in your plan, then go ahead and use the super-short measure.
The Costa and McCrae scales are great. However I recently developed a shorter set of published mini 5 factor scales (see friedmanscales.com) that allows you to assess and score the 5 factors quickly (15 items) and easily give on a weekly (or even daily basis). The Friedman Mini 5 factor scale (2020) is free with norms, SD’s and correlations with many other scales. Check it out
The EVALART 5 Factors Test is an inventory that evaluates the 5 great personality factors: Extraversion, Kindness, Responsibility, Emotional Stability, and Openness to Experience, and is based on the studies of Tupes and Christal (1961) and Norman (1963), who based on the Lexical model of personality, raised the 5 basic characteristics that define personality. The results offer a numerical report on the level that each candidate presents in the characteristics, as well as an exhaustive description of the behaviors , motivations and trends of the same.
The test is sufficiently substantiated at the research level, so it is widely validated when evaluating personality. It also offers a User Manual, which guides the user when carrying out the selection and recruitment process, as it informs about the typical behavioral and emotional patterns that can be found in the different Organizational and Company roles.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers