PSYC 2001 Lecture Notes - Lecture 20: Type I And Type Ii Errors, Null Hypothesis, Statistical Inference

• School: Carleton University
• Department: Psychology
• Course: PSYC 2001
• Professor: Angel Mackenzie

Document Summary

Note: research paper howes and solomon findings. Howes and solomon: correlation of -0. 68 to -0. 75: negative relationship; word frequency and word recognition (duration threshold) In class experiment (conceptual) replication: correlation of 0. 41: positive relationship; word frequency and word recognition (accuracy); testing the same effect why does there appear to be an opposite relationship? (hint: differences in operationalization of the variable (cid:498)word recognition(cid:499)(cid:524) They tested threshold (frequency decreased), but we tested word recognition (frequency increased). We operationalized the variable differently (we measured accuracy) Review of probability and structure of inferential tests. Probability is a key concept underlying inferential statistics. Results of experiments are not viewed as absolute and immutable but instead based on sampling, which is subject to error. Sampling error indicates that means calculated from sample data are estimates of population means. Sampling error is quantifiable size of error bars on a graph. The most common form of inferential statistic uses a procedure known as.