東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

The optimal number of categories for...

McDonald, Jo-Anne L.

Linked to FindBook

Google Book

Amazon

博客來

The optimal number of categories for numerical rating scales.

Record Type:	Electronic resources : Monograph/item
Title/Author:	The optimal number of categories for numerical rating scales./
Author:	McDonald, Jo-Anne L.
Description:	169 p.
Notes:	Source: Dissertation Abstracts International, Volume: 65-05, Section: A, page: 1664.
Contained By:	Dissertation Abstracts International65-05A.
Subject:	Education, Educational Psychology. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3134422
ISBN:	0496816498

The optimal number of categories for numerical rating scales.
McDonald, Jo-Anne L.

The optimal number of categories for numerical rating scales. - 169 p.

Source: Dissertation Abstracts International, Volume: 65-05, Section: A, page: 1664.

Thesis (Ph.D.)--University of Denver, 2004.

This study examined the possibility that an optimal number for numerical rating scales exists. Systematic impact of number of points (NRP) on data quality across Rasch rating scale parameters, internal consistency coefficients, and inter-item correlations was examined using 4-, 5-, 6-, and 7-point Likert-type scales. Participants (N = 394) completed a survey via the World Wide Web that included the Pay Scale Questionnaire, questions concerning preferred NRP, and demographic questions with one of the 4-, 5-, 6-, or 7-point response scales.

ISBN: 0496816498Subjects--Topical Terms:

1017560
Education, Educational Psychology.

The optimal number of categories for numerical rating scales.
LDR:02947nmm 2200313 4500 001 1839891
005 20050714101050.5
008 130614s2004 eng d
020 $a 0496816498
035 $a (UnM)AAI3134422
035 $a AAI3134422
040 $a UnM $c UnM
100 1 $a McDonald, Jo-Anne L. $3 1928255
245 1 4 $a The optimal number of categories for numerical rating scales.
300 $a 169 p.
500 $a Source: Dissertation Abstracts International, Volume: 65-05, Section: A, page: 1664.
500 $a Adviser: Kathy Green.
502 $a Thesis (Ph.D.)--University of Denver, 2004.
520 $a This study examined the possibility that an optimal number for numerical rating scales exists. Systematic impact of number of points (NRP) on data quality across Rasch rating scale parameters, internal consistency coefficients, and inter-item correlations was examined using 4-, 5-, 6-, and 7-point Likert-type scales. Participants (N = 394) completed a survey via the World Wide Web that included the Pay Scale Questionnaire, questions concerning preferred NRP, and demographic questions with one of the 4-, 5-, 6-, or 7-point response scales.
520 $a Internal consistency reliability did not increase significantly (p < .01) as expected beyond the 5-point scale. Average inter-item correlations did not differ across NRP. Person separation indices and item measure estimates did not differ significantly (p < .01) across NRP but differences in model person separation reliability coefficients (MPSR) and step calibrations occurred. The 5- and 6-point rating scales provided high quality data with few malfunctions and MPSRs equivalent to those of the 7-point and higher than the 4-point scales. The 7-point scale was most prone to malfunction and most resistant to improvement. Malfunctions were alleviated in all rating scales by combining categories but MPSR coefficients did not increase as a result of collapsing. The 5-point scale was preferred over the 4-, 6-, and 7-point scales. Reasons for NRP preference included: (a) it covers the range of my opinions, (b) it is easy for me to use, and (c) I prefer to have/not have a middle option.
520 $a In this study, the 5-point rating scale sufficed for responses to the PSQ items, provided for the range of sentiment within this group of participants, elicited high quality data, and was the preferred NRP of the respondents. The researcher recommended that 5-point rating scales be used in situations where moderate responses are expected, that 7-point scales be used when extreme responses are expected, and that 6-point scales be used when a midpoint is an undesirable response option.
590 $a School code: 0061.
650 4 $a Education, Educational Psychology. $3 1017560
650 4 $a Education, Tests and Measurements. $3 1017589
650 4 $a Psychology, Psychometrics. $3 1017742
690 $a 0525
690 $a 0288
690 $a 0632
710 2 0 $a University of Denver. $3 1017461
773 0 $t Dissertation Abstracts International $g 65-05A.
790 1 0 $a Green, Kathy, $e advisor
790 $a 0061
791 $a Ph.D.
792 $a 2004
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3134422