Published: Jan 2009
| ||Format||Pages||Price|| |
|PDF (100K)||2||$25||  ADD TO CART|
|Complete Source PDF (11M)||2||$55||  ADD TO CART|
The mean directional and mean absolute deviations are statistics that can be used to summarize and check the balance of JAR data. The JAR data are analyzed separately for each product and each attribute. The mean directional deviation is the average signed difference of the ratings from the “Just About Right” value. On a 5-point JAR scale, the mean directional ranges from −2 to +2 (“Just Right” =0). Scores that are closer to −2 indicate that respondents thought that attribute was “Too Low.” Scores that are closer to +2 indicate that respondents thought that the attribute was “Too High.” The mean direction deviation is a simple shift of the mean, covered in Appendix D. The mean absolute deviation summarizes the spread of the ratings about the “Just About Right” value. On a 5-point scale, the mean absolute deviation ranges from 0, when all judgments are “Just About Right” to +2, when all judgments are at one or the other extreme end of the scale. Unlike percent “Just-Right” scores, absolute deviations can be calculated for each individual and the mean absolute deviation analyzed by any of the standard parametric statistical procedures. The mean absolute deviation summarizes the average distance to the “Just About Right” value while the mean directional deviation summarizes the average direction the attribute is from the “Just About Right” value (i.e., “Too Low or Too High”).
Fonterra Reserch Center, Palmerston North,
Paper ID: MNL11486M