Variables for Collecting Response Time Data Using Amazon Mechanical Turk Filename: dataExp1.csv Description: Experiment 1 34 adults 48 test items, all ungrammatical wh-questions 72 fillers Presented in randomized order Variables: Count - counter X.1 - counter X - counter fullItem - full item code before it is split response - participant response (only for comprehension questions, 0 = incorrect, 1 = correct) RT - response time, in ms participant - participant ID number item - item number condition - test condition (CondA = singular subject, singular wh, DP; CondB = singular subject, singular wh, pronoun; CondC = plural subject, singular wh, DP; CondD = plural subject, singular wh, pronoun; CondE = singular subject, plural wh, DP; CondF = singular subject, plural wh, pronoun; CondG = plural subject, plural wh, DP; CondF = plural subject, plural wh, pronoun) region Pronoun - no (0) or yes (1) NumberWh - singular (0) or plural (1) NumberSu - singular (0) or plural (1) remove - used to remove items with incorrect comprehension questions word - word read by participant regionTime - response time aggregated by region wordLength - number of characters in word regionLength - number of characters in region logRT - log of region time exp - expected reading time resid - residual reading time Filename: dataExp2.csv Description: Experiment 2 82 participants 16 test items, half ungrammatical 96 fillers Variables: Count - counter X - counter fullItem - full item code before it is split response - participant response (only for comprehension questions, 0 = incorrect, 1 = correct) RT - response time, in ms participant - participant ID number item - item number condition - test condition (CondA = singular attractor, grammatical; CondB = plural attractor, grammatical; CondC = singular attractor, ungrammatical; CondD = singular attractor, ungrammatical) word - word read by participant region - Sentence region in numerical order ungrammatical - no (0) or yes (1) mismatch - number mismatch between subject and attractor, no (0) or yes (1) remove - used to remove items with incorrect comprehension questions wordLength - number of characters in word logRT - log of response time exp - expected reading time resid - residual reading time Filename: dataExp3.csv Description: Experiment 3 60 participants 24 test items, half ungrammatical 72 fillers Count - counter X - counter fullItem - full item code before it is split response - participant response (only for comprehension questions, 0 = incorrect, 1 = correct) RT - response time, in ms participant - participant ID number item - item number condition - test condition (CondA = singular attractor, grammatical; CondB = plural attractor, grammatical; CondC = singular attractor, ungrammatical; CondD = singular attractor, ungrammatical) word - word read by participant region - sentence region in numerical order ungrammatical - no (0) or yes (1) mismatch - number mismatch between subject and attractor, no (0) or yes (1) wordLength - number of characters in word