> &% \pJava Excel API v2.6 Ba==h\:#8X@"1Arial1Arial1Arial1Arial + ) , * `$DC,4"titletitle[alternative]title[translated]author[google(nobreakline)] contributor[author(nobreakline)]"contributor[scopusid(nobreakline)]date[issued]relation[journaltitle]identifier[issn] citationsidentifier[major]subject(nobreakline) publisheridentifier[thesisdegree]contributor[advisor]relation[ispartofseries]relation[index(nobreakline)]typeabstractidentifier[govdoc]identifier[uri]identifier[doi(doi)]identifier[isbn]identifier[ismn]
identifier ɽQǔ 4 0x Likert ĳX 8ĳ DP2002Y ¬YtTŐYP YMasterMaster's ThesisD
There is simply no definitive answer to the question of using a neutral point in response format. In the response format the use of an odd or even number of categories is an issue. An odd number results in neutral response category which allows the respondents to avoid committing to either a positive or negative direction. Proponets of its use suggest that the neutral response is very useful when respondents have no attitude toward an object or have genuinely ambivalent feelings the are equally positive and negative. On the other hand, consistents found that the use of neutral point introduces response styles. This study compares 5 -category scale with 6-category scale to know which one has better validity and reliability.
Th subjects of this research were 296 middle school students and 371 college students and all of them are women. The items of academic self-efficacy were used of the purpose of the present study.
The results from the study are summarized. In term of the neutral response category, the reliability and the validity are not different between 5 category and 6 category when analysis the scales using classical test theory. But 6 category is better on using item response theory. Item information function represents the result of the wide respondents property. Neither of 5 category nor 6 category investigated the method effects.
In conclusion, this research indicated that the use of the neutral point neutral point response category on a Likert Scale is not suggested. 6 category that doesn't take the neutral point shows the better reliability and validity. Therefore, it is suggested to use even number response category on Likert-scale research.; lX @ Likert ¬ĳ ɽQǔ 4 0| ĳX °ĳ@ ĳ (t ǔ| LD, ̹} (t t ɽQǔ| h¤Д JŔ Ő t T @ 8ĳ| | LD t. t| X ɽQǔ h 5 ĳ@ ɽQǔ h J@ 6 ĳX QX| DP X. < P QXX 0 µĬX@ °ĳĬ| X, ٳ|\ xlp| | LDX. 8mQt` p\ 8mX 8ĳ D t ɽQǔ 4 Ǵ (t| tǔ t X. 䲹1-)(MTMM) ,<\ ĳļ 4ĳ ļĳ| LD, 5 ĳ@ 6 ĳX ! ) 0x ) tȬXՔ| LDX.
lX ̸ 0ĳ pXՔ ŐY 371 ŐY 296<\0 . \ QŌ 5 ĳ@ 6 ĳX P XX $8| !X0 5 t 0 L8 p X| .
3. ĳX xlp| LD0 \ Ux xǄ ɽQǔ 4 t ĳX xlpȔ ٳ|X. x iĳ @ 5 ĳ@ 6 ĳ P 买 @ t| ©` ǔ XD .
4. 8mQt`D \ 8mX 8ĳ| XՔp Ǵ 5 ĳ@ 6 ĳX Xǔxļ 8mļĳ \ (tǀ P ĳ X\ (tǰ| ` ̹, 5 ĳ@ 6 ĳX 8mXΨ| 5 ĳ 6 ĳ T @ Q QD Ŵż ǔ ĳD L . 5 ĳX h@ 6 ĳX h| DP\ ĳ 5 ĳ 6 ĳ T @ X ι@ Ō

@
!"#$Root EntryWorkbook*SummaryInformation(DocumentSummaryInformation8