Puzzle Interviews: What Are They and What Do They Measure?

Abstract

Recently organizations have begun using unorthodox interviewing techniques and non-traditional questions to hire applicants. The most notable example of this trend is the so-called interview which was popularized by Microsoft in the 1990s. Puzzle interviews either ask the applicant to solve puzzles (e.g., Why are manhole covers round?) or unusual problems (e.g., How would you weigh an airplane without a scale?). The goals of the present study were a) to determine if a puzzle could be reliably administered and scored, b) to examine whether variation in puzzle performance was related to cognitive ability, and c) to explore the relationships between puzzle performance and applicant reaction measures. Seventy-six undergraduates participated in a videotaped puzzle and completed measures of cognitive ability and applicant reactions. Interview performance ratings by three independent raters yielded an acceptable level of inter-rater reliability. In addition, puzzle performance was significantly correlated with cognitive ability and applicant reaction measures of procedural justice and perceptions of performance. Implications for the use of puzzle interviews in industry and for future research are discussed. The employment is one of the most popular selection and recruiting methods (Posthuma, Morgeson, & Campion, 2002). Employers use interviews for a variety of purposes, including initial screening of job applicants, measurement of job-related attributes, and assessment of person-organization fit. There is also considerable variability in how employers structure and administer interviews, what types of questions they ask, and how they score them. Decades of research on the validity of the employment show that it can be a useful tool in predicting future job performance, particularly if it is highly structured, designed to ask behavioral or situational questions about job-related characteristics, and scored using behaviorally anchored rating scales (Taylor & Small, 2002). Recently, however, some organizations have begun using unorthodox interviewing techniques and non-traditional questions to hire applicants. The most notable example of this trend is the so-called The puzzle was popularized by Microsoft in the 1990s, and is now used in other organizations. According to Poundstone (2003): Puzzle-laden job interviews have infiltrated the Fortune 500 and the rust belt; law firms, banks, consulting firms, and the insurance industry; airlines, media, advertising, and even the armed forces. Brainteaser questions are reported from Italy, Russia, and India. Like it or not, puzzles and riddles are a hot new trend in hiring. (p. 7) Puzzle interviews are generally administered in an unstructured format and do not use formal scoring keys. The most common types of questions either ask the applicant to solve puzzles or brainteasers (e.g., Why are manhole covers round? or How many times a day do a clock's hands overlap?) or to solve unusual problems (e.g., How would you weigh an airplane without a scale? or How many golf balls does it take to fill an Olympic-sized swimming pool?). The conventional wisdom among employers is that these questions measure a variety of applicant characteristics, including critical thinking, creativity, intuition, flexibility, and ability to reason (Munk & Oliver, 1997). In our search of the literature, however, we could find no research evidence to support the use of non-traditional questions in the employment interview. Because of this lack of research, the goals of our study were, a) to determine if a puzzle could be reliably administered and scored and, b) to examine whether variation in puzzle performance could be accounted for by cognitive ability and perceptions of fairness. Developing the Puzzle Interview That we know of, there is no previous research that has examined the psychometric properties of puzzle interviews. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 15 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.