Limits...
Analytical strategies in human growth research.

Johnson W - Am. J. Hum. Biol. (2014)

Bottom Line: A summary table linking each analytical strategy to its applications is provided to help investigators match their hypotheses and measurement schedules to an analysis plan.All too often, serial measurements are treated as cross-sectional in analyses that do not harness the power of longitudinal data.The broad goal of this article is to encourage the rigorous application of longitudinal statistical methods to human growth research.

View Article: PubMed Central - PubMed

Affiliation: MRC Unit for Lifelong Health and Ageing at UCL, London, WC1B 5JU, United Kingdom.

Show MeSH

Related in: MedlinePlus

Example results from growth mixture models with two (A), three (B), four (C), or five (D) latent classes applied to serial body mass index (BMI) data on 417 girls aged 10 to 18 years in the Fels Longitudinal Study.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4309180&req=5

fig03: Example results from growth mixture models with two (A), three (B), four (C), or five (D) latent classes applied to serial body mass index (BMI) data on 417 girls aged 10 to 18 years in the Fels Longitudinal Study.

Mentions: The sample average growth curves of the different models are shown in Figure 3. The model with two latent classes in Panel A identified a “normal weight” class to which 94% of the girls were most likely to belong and an “obese” class to which 6% of the girls were most likely to belong. These classes were present in all subsequent models, with the addition of a “overweight/ pubertal obese” class in Panel B, the further addition of a “overweight/ pre-pubertal obese” class in Panel C, and the even further addition of an “overweight” class in Panel D. Growth mixture models compute probabilities for each participant of belonging to each class; entropy was high for each model (0.911–0.955), thereby indicating that the average probabilities of girls belonging to the class to which they were assigned (based on the one with the highest probability) were close to one. The BIC decreased across the models, from 14,844 in the model with two classes to 14,774 in the model with five classes, thereby indicating that there was marginal improvement in model fit as the number of classes increased.


Analytical strategies in human growth research.

Johnson W - Am. J. Hum. Biol. (2014)

Example results from growth mixture models with two (A), three (B), four (C), or five (D) latent classes applied to serial body mass index (BMI) data on 417 girls aged 10 to 18 years in the Fels Longitudinal Study.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4309180&req=5

fig03: Example results from growth mixture models with two (A), three (B), four (C), or five (D) latent classes applied to serial body mass index (BMI) data on 417 girls aged 10 to 18 years in the Fels Longitudinal Study.
Mentions: The sample average growth curves of the different models are shown in Figure 3. The model with two latent classes in Panel A identified a “normal weight” class to which 94% of the girls were most likely to belong and an “obese” class to which 6% of the girls were most likely to belong. These classes were present in all subsequent models, with the addition of a “overweight/ pubertal obese” class in Panel B, the further addition of a “overweight/ pre-pubertal obese” class in Panel C, and the even further addition of an “overweight” class in Panel D. Growth mixture models compute probabilities for each participant of belonging to each class; entropy was high for each model (0.911–0.955), thereby indicating that the average probabilities of girls belonging to the class to which they were assigned (based on the one with the highest probability) were close to one. The BIC decreased across the models, from 14,844 in the model with two classes to 14,774 in the model with five classes, thereby indicating that there was marginal improvement in model fit as the number of classes increased.

Bottom Line: A summary table linking each analytical strategy to its applications is provided to help investigators match their hypotheses and measurement schedules to an analysis plan.All too often, serial measurements are treated as cross-sectional in analyses that do not harness the power of longitudinal data.The broad goal of this article is to encourage the rigorous application of longitudinal statistical methods to human growth research.

View Article: PubMed Central - PubMed

Affiliation: MRC Unit for Lifelong Health and Ageing at UCL, London, WC1B 5JU, United Kingdom.

Show MeSH
Related in: MedlinePlus