33  Instruments

(Modified on 2026-08-11)
1,073 words

In human-participant research, it is common to make use of standardized tests or measures called instruments. Common instruments include the Wechsler Adult Intelligence Scale (WAIS), the Scholastic Aptitude Test (SAT), and the Minnesota Multiphasic Personality Inventory (MMPI). An instrument often takes the form of a questionnaire, but there are many instruments that involve tests of various sorts, including physical tests. For example, people aiming to become firefighters are typically required to pass a series of tests, including a Firefighter Physical Fitness Test (FPFT); This test includes such tasks as carrying a 75 pound weight up five floors and using a 9-pound sledgehammer to drive a 160-pound steel beam on a Keiser Force Machine.

Literally hundreds of instruments have been developed to measure a wide variety of concepts. These include such tests as the Kinsey Scale of Sexual Orientation (KSRS), the Biddle Agility Test, the Authentic Happiness Inventory Questionnaire, the World Health Organization Quality of Life Questionnaire (WHOQOL), and many many more. In particular, psychologists have created a huge number of instruments intended to measure a wide variety perceptual and cognitive phenomenon: For example, the Beck Depression Inventory (Beck, Steer, and Brown 1996) is intended to characterize how depressed a person feels. The most famous psychology instrument is the so-called “Big 5” personality test (Goldberg 1992), which characterizes a person’s personality in terms of five dimensions: conscientiousness, agreeableness, neuroticism, openness to experience, and extraversion.

There are two broad types of instruments:

Validity and Reliability

Like any empirical measurement, a useful instrument must be both reliable and valid. Most published instruments have been extensively tested for reliability and validity, with multiple peer-reviewed papers to back them up. This frees us from the burden of inventing and justifying our own operational measurement protocols. If you choose to use an existent instrument, you should cite the original paper(s). You might also cite other papers that use the instrument for similar purposes to you.

The first test you ever underwent was probably the Apgar test. All newborns are given an Apgar score between 0 and 10 according to color, pulse, reflex responsiveness, muscle tone, and breathing. Normal scores are in the range of 7 and above. A score of 4 to 6 is considered low, and lower scores indicate a critical condition. The aim of the Apgar test is to provide a quick index of the newborn’s health, and a way of determining whether immediate medical care is needed. The test is quick and easy to do, and is highly predictive of the general medical health of the infant. That is, the test has a reasonably good validity.

This last point brings to mind another important criterion: ease! In most human-participant research, you have limited time with participants. Being able to take a measurement quickly is very helpful. All instruments, and measurement protocols in general, must trade-off between speed/ease and validity/reliability. Quick and simple instruments may be less valid but still be better to use overall if it means you have more time to ask other questions.

Musical “Instruments”

Though many psychological instruments have been developed and published, there are only a handful of music specific instruments which get much use. One is the Montreal Battery of Evaluation of Amusia (MBEA) developed by Isabelle Peretz (2003) and her colleagues. This instrument includes a number of listening and production tasks that can be used to assess “tone-deafness.” Another is the Geneva Emotional Music Scale (GEMS), which is used to measure emotions evoked by music (Zentner, Grandjean, and Scherer 2008). There are actually several GEMS—GEMS-9, GEMS-25, and GEMS-45—with the numbers indicating the number of questions on the scale. The smaller instrument (GEMS-9) can be filled out more rapidly, but the larger scales (GEMS-45) give more nuanced emotional information.

Musical Sophistication

In music-related experiments, it is common for researchers to distinguish “musicians” from “non-musicians.” Of course, there is no clear dividing line between musicians and non-musicians, but it is often useful to operationally define this line in order to pursue research. Two of the most commonly used criteria are “number of years of private music lessons” and “music major versus non-music major.” However, given the huge impact that musical ability has on music perception, many researchers have sought more nuanced, reliable, and valid ways of measuring “musicianness.”

The earliest instrument for measuring “musicianness” was developed at Ohio State University: the Ollen Musical Sophistication Index (OMSI) (Ollen 2006). This instrument consists of ten questions that are used to estimate a person’s degree of musical sophistication. The Goldsmiths Musical Sophistication Index (Gold-MSI) was developed later (Müllensiefen et al. 2014), and is now much more widely used. The Gold-MSI is a much more elaborate index for assessing musicality and is not limited to a musician population. It involves various perceptual tasks, and doesn’t rely solely on self-report. However, the Gold-MSI takes much longer to administer than the OMSI. In fact, the quickest way to effectively measure “musical sophistication” is using simply a single item from the OMSI (Zhang and Schubert 2019):

Which title best describes you?

  • Nonmusician
  • Music-loving nonmusician
  • Amateur musician
  • Serious amateur musician
  • Semiprofessional musician
  • Professional musician

Developing Instruments

The development of psychological instruments is itself a huge field of research. You could have an entire career as a scientist developing instruments to measure various things! We won’t learn too much about how this is done. Generally, the process starts with initial exploratory, qualitative inquiries, which are used to develop and refine ideas for the instrument. When it comes to perceptual or cognitive instruments, the final step is usually factor analysis—a statistical method for reducing a large number of possible measures into a smaller number which are highly correlated with the full set. For example, when developing GEMS, researchers started by studying over 400 different emotional terms, but used factor analysis to reduce these terms to the 45-, 25-, and 9-dimensional scales that we use.

Ollen Musical Sophistication Index

We can get some idea of how instruments are developed, by considering the development of the Ollen Musical Sophistication Index (Ollen 2006). Joy Ollen was eager to produce a refined instrument for “musical sophistication” that could be quickly administered in the context of typical music-related research. Here are the steps she followed:

  1. Ollen first carried out a literature review—identifying the kinds of criteria used by researchers for distinguishing the level of musical sophistication. She identified 743 published music studies and found 38 different indicators used to determine degree of musicality.
  2. Ollen then carried out a survey of 27 experienced musicians and asked them what sorts of questions they might ask to determine how musically sophisticated an individual might be. These musicians identified many of the same 38 criteria found in the published studies, but they also added four new criteria.
  3. Ollen distilled the 42 criteria to the most promising 29 criteria. She then developed a 36-item questionnaire which asked such questions as the number of years of private musical instruction, whether the individual had taken college-level music courses, the number of hours per week in musical practice, and so on.
  4. She then administered this questionnaire to 633 adults who participated in various groups involved in music-related activities. This ranged from members of church choirs to members of professional orchestras. The leader of each group (a musical expert) was then asked to rate the musical sophistication of every member of the group using a 10-point scale.
  5. The data was then analyzed to determine which of the 36 questions were most predictive of the experts’ assessments of musical sophistication. Using a logistic regression analysis, nine indicators were identified that accounted for nearly 80% of the variability. From this, Ollen created a 10-item questionnaire that can be completed in under 2 minutes.
  6. Finally, Ollen administered her instrument to a large independent sample of people. Once again, group leaders rated the musicality of the group members. Although the accounted variance dropped to about 50%, the results show that the OMSI instrument provides reasonably valid estimates of musical sophistication with a minimum of effort.