Intertemporal Differences Among MTurk Workers: Time-Based Sample Variations and Implications for Online Data Collection:

SAGE Open

Logan S. Casey¹ 

关键词: political methodology;    political science;    social sciences;    politics;    social sciences;    research methods;    data collection;    research methodology;    design;    reliability;    validity;    political behavior/psychology;    psychology;   

DOI  :  10.1177/2158244017712774

学科分类：社会科学、人文和艺术（综合）

来源: Sage Journals

【 摘 要 】

The online labor market Amazon Mechanical Turk (MTurk) is an increasingly popular source of respondents for social science research. A growing body of research has examined the demographic composition of MTurk workers as compared with that of other populations. While these comparisons have revealed the ways in which MTurk workers are and are not representative of the general population, variations among samples drawn from MTurk have received less attention. This article focuses on whether MTurk sample composition varies as a function of time. Specifically, we examine whether demographic characteristics vary by (a) time of day, (b) day of week, and serial position (i.e., earlier or later in data collection), both (c) across the entire data collection and (d) within specific batches. We find that day of week differences are minimal, but that time of day and serial position are associated with small but important variations in demographic composition. This demonstrates that MTurk samples cannot be presumed identical across different studies, potentially affecting reliability, validity, and efforts to reproduce findings.

【 授权许可】

CC BY