The Alzheimer's Disease Assessment Scale - Cognitive ...



Appendix 2: Psychometric properties, their definitions and criteria for satisfactory performance Psychometric propertyDefinitionTest(s)Criteria for AcceptabilityData CompletenessThe extent to which ADAS-cog components are scored and ADAS-cog total scores can be computed. Computing the percent of missing data for each component, and the percent of people for whom a scale score can be computed.[ ADDIN EN.CITE <EndNote><Cite><Author>McHorney</Author><Year>1994</Year><RecNum>592</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>McHorney, C.A.</AUTHOR><AUTHOR>Ware, J.E.Jr.</AUTHOR><AUTHOR>Lu, J.F.R.</AUTHOR><AUTHOR>Sherbourne, C.D.</AUTHOR></AUTHORS><YEAR>1994</YEAR><TITLE>The MOS 36-Item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions and reliability across diverse patient groups</TITLE><SECONDARY_TITLE>Medical Care</SECONDARY_TITLE><VOLUME>32</VOLUME><NUMBER>1</NUMBER><PAGES>40-66</PAGES></MDL></Cite></EndNote>27] Component-level missing data is <10%[ ADDIN EN.CITE <EndNote><Cite><Author>Group</Author><Year>1998</Year><RecNum>333</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>The WHOQOL Group</AUTHOR></AUTHORS><YEAR>1998</YEAR><TITLE>The World Health Organisation Quality of Life Assessment (WHOQOL): development and general psychometric properties</TITLE><SECONDARY_TITLE>Social Science and Medicine</SECONDARY_TITLE><VOLUME>46</VOLUME><NUMBER>12</NUMBER><PAGES>1569-1585</PAGES></MDL></Cite></EndNote>28] Computable scale scores >50% completed components.[ ADDIN EN.CITE <EndNote><Cite><Author>Ware</Author><Year>1993</Year><RecNum>854</RecNum><MDL><REFERENCE_TYPE>1</REFERENCE_TYPE><AUTHORS><AUTHOR>Ware, J.E. Jr.</AUTHOR><AUTHOR>Snow, K.K.</AUTHOR><AUTHOR>Kosinski, M.</AUTHOR><AUTHOR>Gandek, B.</AUTHOR></AUTHORS><YEAR>1993</YEAR><TITLE>SF-36 Health Survey manual and interpretation guide</TITLE><PLACE_PUBLISHED>Boston, Massachusetts</PLACE_PUBLISHED><PUBLISHER>Nimrod Press</PUBLISHER></MDL></Cite></EndNote>29] A total score can only be computed if all components are scored as they have substantially different ranges.Scaling AssumptionsThe extent to which it is legitimate to sum a set of component scores, without weighting or standardisation, to produce a single total score.[ ADDIN EN.CITE <EndNote><Cite><Author>Hays</Author><Year>1990</Year><RecNum>380</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Hays, R.D.</AUTHOR><AUTHOR>Hayashi, T.</AUTHOR></AUTHORS><YEAR>1990</YEAR><TITLE>Beyond internal consistency reliability: rationale and user&apos;s guide for Multi-Trait Analysis Program on the microcomputer</TITLE><SECONDARY_TITLE>Behavior Research Methods, Instruments, &amp; Computers</SECONDARY_TITLE><VOLUME>22</VOLUME><PAGES>167-175</PAGES></MDL></Cite></EndNote>30 ADDIN EN.CITE <EndNote><Cite><Author>DeVellis</Author><Year>1991</Year><RecNum>188</RecNum><MDL><REFERENCE_TYPE>1</REFERENCE_TYPE><AUTHORS><AUTHOR>DeVellis, R.F.</AUTHOR></AUTHORS><YEAR>1991</YEAR><TITLE>Scale development: theory and applications</TITLE><SECONDARY_TITLE>Applied social research methods</SECONDARY_TITLE><PLACE_PUBLISHED>London</PLACE_PUBLISHED><PUBLISHER>Sage publications</PUBLISHER><VOLUME>26</VOLUME><NUMBER_OF_VOLUMES>28</NUMBER_OF_VOLUMES><PAGES>121</PAGES><KEYWORDS><KEYWORD>Scale developement</KEYWORD><KEYWORD>Psychometric evaluation</KEYWORD><KEYWORD>Reliability</KEYWORD><KEYWORD>Validity</KEYWORD><KEYWORD>Measurement</KEYWORD><KEYWORD>Factor analysis</KEYWORD></KEYWORDS></MDL></Cite></EndNote>31]Summing ADAS-cog component scores is considered legitimate, when the components:Are approximately parallell ( ie they measure at the same point on the scale). Contribute similarly to the variation of the total score (ie they have similar variances), otherwise these should be standardised. Measure a common underlying construct (ie cognitive performance)[ ADDIN EN.CITE <EndNote><Cite><Author>Guttman</Author><Year>1954</Year><RecNum>346</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Guttman, L.A.</AUTHOR></AUTHORS><YEAR>1954</YEAR><TITLE>Some necessary conditions for common-factor analysis</TITLE><SECONDARY_TITLE>Psychometrika</SECONDARY_TITLE><VOLUME>19</VOLUME><NUMBER>2</NUMBER><PAGES>149-161</PAGES></MDL></Cite></EndNote>32] otherwise combining them to produce a single score is not appropriate. Contain a similar proportion of information concerning the construct being measured. Otherwise components should be given different weights.[ ADDIN EN.CITE <EndNote><Cite><Author>Likert</Author><Year>1932</Year><RecNum>545</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Likert, R.A.</AUTHOR></AUTHORS><YEAR>1932</YEAR><TITLE>A technique for the measurement of attitudes</TITLE><SECONDARY_TITLE>Archives of Psychology</SECONDARY_TITLE><VOLUME>140</VOLUME><PAGES>5-55</PAGES></MDL></Cite></EndNote>33] Satisfied when components have similar mean scores.[ ADDIN EN.CITE <EndNote><Cite><Author>Likert</Author><Year>1932</Year><RecNum>545</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Likert, R.A.</AUTHOR></AUTHORS><YEAR>1932</YEAR><TITLE>A technique for the measurement of attitudes</TITLE><SECONDARY_TITLE>Archives of Psychology</SECONDARY_TITLE><VOLUME>140</VOLUME><PAGES>5-55</PAGES></MDL></Cite></EndNote>33]Satisfied when components have similar standard deviations.[ ADDIN EN.CITE <EndNote><Cite><Author>McHorney</Author><Year>1994</Year><RecNum>592</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>McHorney, C.A.</AUTHOR><AUTHOR>Ware, J.E.Jr.</AUTHOR><AUTHOR>Lu, J.F.R.</AUTHOR><AUTHOR>Sherbourne, C.D.</AUTHOR></AUTHORS><YEAR>1994</YEAR><TITLE>The MOS 36-Item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions and reliability across diverse patient groups</TITLE><SECONDARY_TITLE>Medical Care</SECONDARY_TITLE><VOLUME>32</VOLUME><NUMBER>1</NUMBER><PAGES>40-66</PAGES></MDL></Cite></EndNote>27]Satisfied when components have adequate corrected component-total correlation (ITC 0.30).[ ADDIN EN.CITE <EndNote><Cite><Author>Ware</Author><Year>1997</Year><RecNum>863</RecNum><MDL><REFERENCE_TYPE>1</REFERENCE_TYPE><AUTHORS><AUTHOR>Ware, J.E.Jr.</AUTHOR><AUTHOR>Harris, W.J.</AUTHOR><AUTHOR>Gandek, B.</AUTHOR><AUTHOR>Rogers, B.W.</AUTHOR><AUTHOR>Reese, P.R.</AUTHOR></AUTHORS><YEAR>1997</YEAR><TITLE>MAP-R for windows: multitrait / multi-item analysis program - revised user&apos;s guide.</TITLE><PLACE_PUBLISHED>Boston, MA</PLACE_PUBLISHED><PUBLISHER>Health Assessment Lab.</PUBLISHER></MDL></Cite></EndNote>34]Satisfied when components have similar ITCs.[ ADDIN EN.CITE <EndNote><Cite><Author>Ware</Author><Year>1997</Year><RecNum>863</RecNum><MDL><REFERENCE_TYPE>1</REFERENCE_TYPE><AUTHORS><AUTHOR>Ware, J.E.Jr.</AUTHOR><AUTHOR>Harris, W.J.</AUTHOR><AUTHOR>Gandek, B.</AUTHOR><AUTHOR>Rogers, B.W.</AUTHOR><AUTHOR>Reese, P.R.</AUTHOR></AUTHORS><YEAR>1997</YEAR><TITLE>MAP-R for windows: multitrait / multi-item analysis program - revised user&apos;s guide.</TITLE><PLACE_PUBLISHED>Boston, MA</PLACE_PUBLISHED><PUBLISHER>Health Assessment Lab.</PUBLISHER></MDL></Cite></EndNote>34]TargetingThe extent to which the range of the variable measured by the scale (here cognitive performance) matches the range of that variable in the study sample. Score distributions were examined at both the ADAS-cog component and scale level. This was conducted in the whole sample and in AD severity subgroups defined by three MMSE ranges: 10-14 (marked); 15-20 (moderate); 21-26 (mild).Scale scores should span the entire range; floor (proportion of the sample at the maximum scale score for the ADAS-cog) and ceiling (proportion of the sample at the minimum scale score) effects should be low (<15%);[ ADDIN EN.CITE <EndNote><Cite><Author>McHorney</Author><Year>1995</Year><RecNum>594</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>McHorney, C.A.</AUTHOR><AUTHOR>Tarlov, A.R.</AUTHOR></AUTHORS><YEAR>1995</YEAR><TITLE>Individual-patient monitoring in clinical practice: are available health status surveys adequate?</TITLE><SECONDARY_TITLE>Quality of Life Research</SECONDARY_TITLE><VOLUME>4</VOLUME><PAGES>293-307</PAGES><KEYWORDS><KEYWORD>Floor and ceiling effects</KEYWORD><KEYWORD>Standard error of measurement</KEYWORD><KEYWORD>score distributions</KEYWORD><KEYWORD>Validity</KEYWORD><KEYWORD>Reliability</KEYWORD></KEYWORDS></MDL></Cite></EndNote>35] and skewness statistics should range from –1 to +1.[ ADDIN EN.CITE <EndNote><Cite><Author>Hays</Author><Year>1993</Year><RecNum>383</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Hays, R.D.</AUTHOR><AUTHOR>Anderson, R.</AUTHOR><AUTHOR>Revicki, D.A.</AUTHOR></AUTHORS><YEAR>1993</YEAR><TITLE>Psychometric considerations in evaluating health-related quality of life measures</TITLE><SECONDARY_TITLE>Quality of Life Research</SECONDARY_TITLE><VOLUME>2</VOLUME><PAGES>441-449</PAGES></MDL></Cite></EndNote>36]There is no published criteria for component level targeting. Therefore, we applied the scale-level criteria. This is frequently overlooked but important.[ ADDIN EN.CITE <EndNote><Cite><Author>O&apos;Connor</Author><Year>2004</Year><RecNum>1284</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>O&apos;Connor, R.J.</AUTHOR><AUTHOR>Cano, S.J.</AUTHOR><AUTHOR>Thompson, A.J.</AUTHOR><AUTHOR>Hobart, J.C.</AUTHOR></AUTHORS><YEAR>2004</YEAR><TITLE>Exploring rating scale responsiveness: does the total score reflect the sum of its parts?</TITLE><SECONDARY_TITLE>Neurology</SECONDARY_TITLE><VOLUME>62</VOLUME><PAGES>1842-1844</PAGES></MDL></Cite></EndNote>37]ReliabilityReliability is the extent to which scale scores are associated with random error. High reliability indicates that scores are associated with little random error, i.e. are consistent. Two types of reliability were examined at both scale and component level. Each quantifies a different source of random error: Internal consistency reliability estimates the random error associated with total scores from the intercorrelations among the components.[ ADDIN EN.CITE <EndNote><Cite><Author>Cronbach</Author><Year>1951</Year><RecNum>159</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Cronbach, L.J.</AUTHOR></AUTHORS><YEAR>1951</YEAR><TITLE>Coefficient alpha and the internal structure of tests</TITLE><SECONDARY_TITLE>Psychometrika</SECONDARY_TITLE><VOLUME>16</VOLUME><NUMBER>3</NUMBER><PAGES>297-334</PAGES><KEYWORDS><KEYWORD>JNNP Editorial: measuring outcome-the bare essentials.</KEYWORD></KEYWORDS></MDL></Cite></EndNote>38] Test retest (TRT) reproducibility, based on the agreement between people scores at screening and baseline, estimates the ability of components and scales to produce stable scores.[ ADDIN EN.CITE <EndNote><Cite><Author>Hays</Author><Year>1993</Year><RecNum>383</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Hays, R.D.</AUTHOR><AUTHOR>Anderson, R.</AUTHOR><AUTHOR>Revicki, D.A.</AUTHOR></AUTHORS><YEAR>1993</YEAR><TITLE>Psychometric considerations in evaluating health-related quality of life measures</TITLE><SECONDARY_TITLE>Quality of Life Research</SECONDARY_TITLE><VOLUME>2</VOLUME><PAGES>441-449</PAGES></MDL></Cite></EndNote>36] Recommended for adequate scale internal consistency is Cronbach's alpha coefficient 0.80,[ ADDIN EN.CITE <EndNote><Cite><Author>Cronbach</Author><Year>1951</Year><RecNum>159</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Cronbach, L.J.</AUTHOR></AUTHORS><YEAR>1951</YEAR><TITLE>Coefficient alpha and the internal structure of tests</TITLE><SECONDARY_TITLE>Psychometrika</SECONDARY_TITLE><VOLUME>16</VOLUME><NUMBER>3</NUMBER><PAGES>297-334</PAGES><KEYWORDS><KEYWORD>JNNP Editorial: measuring outcome-the bare essentials.</KEYWORD></KEYWORDS></MDL></Cite></EndNote>38] and item internal consistency is item total correlations >0.40.Recommended for adequate TRT reproducibility are scale-level intraclass correlation coefficients (ICC) 0.80[ ADDIN EN.CITE <EndNote><Cite><Author>Nunnally</Author><Year>1994</Year><RecNum>654</RecNum><MDL><REFERENCE_TYPE>1</REFERENCE_TYPE><AUTHORS><AUTHOR>Nunnally, Jum C</AUTHOR><AUTHOR>Bernstein, Ira H</AUTHOR></AUTHORS><YEAR>1994</YEAR><TITLE>Psychometric theory</TITLE><PLACE_PUBLISHED>New York</PLACE_PUBLISHED><PUBLISHER>McGraw-Hill</PUBLISHER><EDITION>3rd</EDITION></MDL></Cite></EndNote>39] and item level ICC 0.50.[ ADDIN EN.CITE <EndNote><Cite><Author>Duruoz</Author><Year>1996</Year><RecNum>1038</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Duruoz, M.T.</AUTHOR><AUTHOR>Poiraudeau, S.</AUTHOR><AUTHOR>Fermanian, J.</AUTHOR><AUTHOR>Menkes, C-J.</AUTHOR><AUTHOR>Amor, B.</AUTHOR><AUTHOR>Dougados, M.</AUTHOR><AUTHOR>Revel, M.</AUTHOR></AUTHORS><YEAR>1996</YEAR><TITLE>Development and validation of a rheumatoid hand functional disability scale that assesses functional handicap</TITLE><SECONDARY_TITLE>Journal of Rheumatology</SECONDARY_TITLE><VOLUME>23</VOLUME><PAGES>1167-1172</PAGES></MDL></Cite></EndNote>40]ValidityThe extent to which a scale measures what it intends to measure and is essential for the accurate and meaningful interpretation of scores.[ ADDIN EN.CITE <EndNote><Cite><Author>Kaplan</Author><Year>1976</Year><RecNum>466</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Kaplan, R.M.</AUTHOR><AUTHOR>Bush, J.W.</AUTHOR><AUTHOR>Barry, C.C.</AUTHOR></AUTHORS><YEAR>1976</YEAR><TITLE>Health status: types of validity and the index of well-being</TITLE><SECONDARY_TITLE>Health Services Research</SECONDARY_TITLE><VOLUME>11</VOLUME><NUMBER>Winter</NUMBER><PAGES>478-507</PAGES></MDL></Cite></EndNote>41] Three aspects of construct validity were tested:Convergent construct validity was examined by computing correlations between ADAS-cog and the mini mental state examination (MMSE[ ADDIN EN.CITE <EndNote><Cite><Author>Folstein</Author><Year>1975</Year><RecNum>259</RecNum><MDL><REFERENCE_TYPE>0</REFERENCE_TYPE><AUTHORS><AUTHOR>Folstein, M.F.</AUTHOR><AUTHOR>Folstein, S.E.</AUTHOR><AUTHOR>McHugh, P.R.</AUTHOR></AUTHORS><YEAR>1975</YEAR><TITLE>&quot;Mini-Mental State&quot;: a practical method for grading the cognitive state of patients for the clinician</TITLE><SECONDARY_TITLE>Journal of Psychiatric Research</SECONDARY_TITLE><VOLUME>12</VOLUME><PAGES>189-198</PAGES><KEYWORDS><KEYWORD>Measurement</KEYWORD><KEYWORD>Cognitive state</KEYWORD></KEYWORDS></MDL></Cite></EndNote>42]). Discriminant construct validity was examined by computing correlations between the ADAS-cog and sociodemographic variables (age and sex) to determine the extent to which they were biased by these variables. Group difference construct validity was examined by comparing ADAS-cog mean scores for the three MMSE defined groups.We hypothesised that the ADAS-cog and MMSE would be highly negatively correlated (r>-0.70) as the two scales measure cognitive performance but are scored in opposite directions.We predicted ADAS-cog scores should not be notably biased by these variables and, therefore, correlations would be low <0.30.We predicted a stepwise change in ADAS-cog scores across the groups, and that the means scores would be signifincatly different. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download