Measurement theory Flashcards
Terms : Hide Images [1]
| 2731402990 | Types of differentiation | Nominal (name) Numeral (# assignment w/o meaning) Number | 0 | |
| 2731402991 | Measurement scales | Nominal: "one or the other" (Y/N, M/F) Ordinal: categories that don't have equidistant measures than them (MMT) Interval: order and distance w/o an absolute zero (temp, months of the year) Ratio: origin, distance and order (gait speed, days since injury) | 1 | |
| 2731402992 | Discreet measurement | Nominal, distinct categories | 2 | |
| 2731402993 | Continuous measurement | Values limited by measurement Taken with different degrees (sec v. msec) | 3 | |
| 2731402994 | Frequency distribution | Number of times something appears in a data set | 4 | |
| 2731402995 | Normal distribution | 68% of data falls w/in 1 standard deviation of the mean | 5 | |
| 2731402996 | Positive skew | To the left Only a few people performed well | 6 | |
| 2731402997 | Negative skew | To the right Only a few people performed poorly | 7 | |
| 2731402998 | Variance | Variability around the mean | 8 | |
| 2731402999 | When is a small variance warranted? Large variance? | Small for scores close together Large for scores far apart | 9 | |
| 2731403000 | Standard deviation | Bringing the variance back into same measurement as the mean | 10 | |
| 2731403001 | Indication of a small standard deviation? Large? | Small: group is homogeneous, more indicative that change has occurred Large: group is heterogeneous, harder to show change in population | 11 | |
| 2731403002 | Correlation coefficient | The relationship between two things 0= none, 1=perfect | 12 | |
| 2731403003 | Positive v. Negative correlation | Positive: direct relationship Negative: indirect relationship | 13 | |
| 2731403004 | Standard error of measurement | Reliability of response when data is normally distributed (The differentiation between true change and error) | 14 | |
| 2731403005 | Norm reference standard | Compare scores of individuals | 15 | |
| 2731403006 | Criterion reference standard | Comparison of an individual to a standard | 16 | |
| 2731455378 | Relative reliability | The degree with which measurements maintain a relationship when measured repeatedly Measured with ICC | 17 | |
| 2731455379 | Absolute reliability | The variability of an individual score measured repeatedly Measured with SEM | 18 | |
| 2731455380 | High or low: What type of ICC do you want? SEM? | High ICC Low SEM | 19 | |
| 2731455381 | Content validity | The extent to which a measure accurately represents a variable of interest | 20 | |
| 2731455382 | Floor effect | Pt. is declining, but can't score any lower | 21 | |
| 2731455383 | Ceiling effect | Pt. is improving, but can't score any higher | 22 | |
| 2731455384 | Responsiveness | How well does the tool measure the change? | 23 | |
| 2731455385 | MDC v. MCID | MCD: Smallest amnt. of "true change" MCID: Smallest amnt. of MEANINGFUL change | 24 | |
| 2731455386 | MCD or MCID, which is more critical for assessing the benefit of Txt? | MCID | 25 |