Who Are the Turkers? A Characterization of MTurk Workers Using the Personality Assessment Inventory

Abstract

As online data collection services such as Amazon's Mechanical Turk (MTurk) gain popularity, the quality and representativeness of such data sources have gained research attention. To date, the majority of existing studies have compared MTurk workers with undergraduate samples, localized community samples, or other Internet-based samples, and thus, there remains little known about the personality and mental health constructs of MTurk workers relative to a national representative sample. The present study addresses these limitations and broadens the scope of existing research through the use of the Personality Assessment Inventory, a multiscale, self-report questionnaire which provides information regarding data validity and personality and psychopathology features standardized against a national U.S. census-matched normative sample. Results indicate that MTurk workers generally provide high-quality data and are reasonably representative of the general population across most psychological dimensions assessed. However, several distinguishing features of MTurk workers emerged that were consistent with prior findings of such individuals, primarily involving somewhat higher negative affect and lower social engagement.

Convergence between Rorschach and self‐report: A new look at some old questions

“…Future research in clinical samples, particularly those demonstrating psychotic processes and thus manifesting the full range of such constructs, will hopefully better illuminate the nature of these relationships. Nonetheless, there is evidence to suggest that MTurk workers do exhibit considerable variability with respect to most psychopathological constructs, and may be somewhat above average in areas of social detachment and negative affect (McCredie & Morey, ). As such, this sample may be more representative of clinical phenomena than other frequently used nonclinical samples, such as undergraduates.…”

Food addiction among men and women in India

“…Comparisons of the psychometric properties of measures completed by Mechanical Turk participants have similar reliability and validity as those completed by participants recruited using traditional sources (Hauser & Schwarz, ; Shapiro, Chandler, & Mueller, ). Mechanical Turk has increasing popularity in the social sciences (Buhrmester et al, ) and has been used in psychological and psychiatric research (Mathes, Norr, Allan, Albanese, & Schmidt, ; McCredie & Morey, ), including eating research (Burrows et al, ; Powell, Frankel, Umemura, & Hazen, ). For the present investigation, participants were paid 50 cents, consistent with most Mechanical Turk research (Buhrmester et al, ).…”

“…Individuals were eligible to participate and included in study analyses if they were over 21 years old, spoke English, and were born and currently resided in India. Mechanical Turk is a recruitment platform that yields high-quality and convenient participant data (Buhrmester, Kwang, & Gosling, 2011;Buhrmester, Talaifar, & Gosling, 2018;McCredie & Morey, 2018), in part by including checks for attentiveness during survey completion. Samples have greater demographic (e.g., age) and geographical diversity than traditionally recruited convenience samples, such as undergraduate students and other internet samples (Behrend, Sharek, Meade, & Wiebe, 2011) but otherwise have similar characteristics to samples recruited by other methodologies (McCredie & Morey, 2018).…”

Convergence between Thematic Apperception Test (TAT) and self‐report: Another look at some old questions

“…As also observed with the Amplified Multiple Choice Test (McCredie & Morey, ), the vast number of significant interrelationships raises potential issues of discriminant validity that warrant further investigation. Although hypothesized correlations were observed in the expected directions, the additional number of medium to large effect size correlations across nearly all of the PAI and IPIP‐NEO‐50 indicators suggests that these scoring categories may be limited in their ability to detect discrete psychological or personality features.…”

“…Response alternatives range from “totally false, not at all true” (1) to “very true” (4). Means, SDs, and internal consistencies of the PAI scales and subscales in the present sample are published in a previous study (McCredie & Morey, ). Internal consistency alpha coefficients ranged from .81 (Stress) to .95 (ANX) for the full scales and .63 (obsessive–compulsive; ARD‐O) to .91 (affective depression; DEP‐A) for the subscales.…”