HD FS 312W Lecture Notes - Lecture 8: Apache Hadoop, Internal Consistency

• School: PSU
• Department: Human Development and Family Studies
• Course: HD FS 312W
• Professor: Rose Wesche

Document Summary

What limitations do you have: time (spatial, financial. All data results from some form of measurement. True score: your actual score on a measure: ex: actual weight when measured. Measurement error: error in data collection based on human error or poor measurement instruments. Observed score = true score + error (reality) Influenced by: human error, poor measurement instrument. How to minimize measurement error: with reliable instruments. A measure of the consistency or reproducibility of data collected using the same methodology. Test-retest reliability: do you get the same answer/score. Inter-rater reliability: do observers score the same behaviors in the same way. Across different but related test items: split-half reliability: do you get the same answer/score with different combinations of questions in one test. How responses to half of the questions similar to other half: equivalent forms reliability: do you get same answer/score with different forms of test, internal consistency: are all items in the measure highly correlated with other items.