國立虎尾科技大學 |

Using the Multilevel Generalized Mixed Model to Impute Missing Accelermometry.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Using the Multilevel Generalized Mixed Model to Impute Missing Accelermometry./
作者:	Wilson, Rebbecca P.
面頁冊數:	1 online resource (111 pages)
附註:	Source: Dissertation Abstracts International, Volume: 79-10(E), Section: B.
Contained By:	Dissertation Abstracts International79-10B(E).
標題:	Biostatistics. -
電子資源:	click for full text (PQDT)
ISBN:	9780438064898

Using the Multilevel Generalized Mixed Model to Impute Missing Accelermometry.
Wilson, Rebbecca P.

Using the Multilevel Generalized Mixed Model to Impute Missing Accelermometry. - 1 online resource (111 pages)

Source: Dissertation Abstracts International, Volume: 79-10(E), Section: B.

Thesis (Dr.P.H.)--The University of North Carolina at Chapel Hill, 2018.

Includes bibliographical references

Accelerometers provide objective measures of physical activity and sedentary behavior. Typically, the device is worn for one week during all waking hours to measure physical activity counts for a period time (e.g., minute). A challenge is accounting for time when the accelerometer is not worn which can bias assessments of physical activity and sedentary behavior. To circumvent this, researchers will limit analysis to participants with a minimum number of adherent days with sufficient wear time and average these days. Excluding accelerometer nonwear assumes missing completely at random (MCAR); yet, sedentary behavior and physical activity are related to nonwear.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2018

Mode of access: World Wide Web

ISBN: 9780438064898Subjects--Topical Terms:

783654
Biostatistics.
Index Terms--Genre/Form:

554714
Electronic books.

Using the Multilevel Generalized Mixed Model to Impute Missing Accelermometry.
LDR:03394ntm a2200361Ki 4500 001 918888
005 20181106104114.5
006 m o u
007 cr mn||||a|a||
008 190606s2018 xx obm 000 0 eng d
020 $a 9780438064898
035 $a (MiAaPQ)AAI10790311
035 $a (MiAaPQ)unc:17812
035 $a AAI10790311
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Wilson, Rebbecca P. $3 1193330
245 1 0 $a Using the Multilevel Generalized Mixed Model to Impute Missing Accelermometry.
264 0 $c 2018
300 $a 1 online resource (111 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertation Abstracts International, Volume: 79-10(E), Section: B.
500 $a Advisers: Shrikant Bangdiwala; Daniela Sotres-Alvarez.
502 $a Thesis (Dr.P.H.)--The University of North Carolina at Chapel Hill, 2018.
504 $a Includes bibliographical references
520 $a Accelerometers provide objective measures of physical activity and sedentary behavior. Typically, the device is worn for one week during all waking hours to measure physical activity counts for a period time (e.g., minute). A challenge is accounting for time when the accelerometer is not worn which can bias assessments of physical activity and sedentary behavior. To circumvent this, researchers will limit analysis to participants with a minimum number of adherent days with sufficient wear time and average these days. Excluding accelerometer nonwear assumes missing completely at random (MCAR); yet, sedentary behavior and physical activity are related to nonwear.
520 $a We proposed imputing counts/min for nonwear using a multilevel generalized mixed model (MGMM) and account for multivariate counts under a complex survey design. Using data from the Hispanic Community Health Study/ Study of Latinos (2008 -- 2011), and the HCHS/SOL Youth Study (2012 -- 2014), we utilize these methods to: (1) compare accelerometer wear and nonwear data in intervals of the day, (2) determine an association between accelerometer average counts/min and BMI, and (3) evaluate the different models using percent relative bias in simulated data.
520 $a Our results showed that (1) accelerometer average counts/min were higher for wear versus nonwear segments in an interval, thus, we concluded that the MCAR assumption of the ad hoc approach was not tenable. (2) The MGMM indicated a clear association between average count/min and BMI when missing values were imputed at the interval level. (3) The percent relative bias did not show enough evidence to support a smaller value for MGMM imputation evaluation models that were concordant with MGMM generated data. We concluded that imputing missing values at the smallest unit possible (e.g., interval), and then aggregating at the participant level, may reduce the potential for making a type 2 error.
520 $a Further research in this area will greatly improve physical activity guidelines established using accelerometer data that better accounts for nonwear time.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Biostatistics. $3 783654
655 7 $a Electronic books. $2 local $3 554714
690 $a 0308
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a The University of North Carolina at Chapel Hill. $b Biostatistics. $3 1193116
773 0 $t Dissertation Abstracts International $g 79-10B(E).
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10790311 $z click for full text (PQDT)