Garantie de satisfaction à 100% Disponible immédiatement après paiement En ligne et en PDF Tu n'es attaché à rien
logo-home
Recherché précédemment par vous
Recherché précédemment par vous
logo
VU Master Health Sciences: Advanced Statistics - Summary lectures and literature: How to analyse clustered and longitudinal data and how to interpret the results? 7,39 €
Ajouter au panier
Accédez rapidement à
Aperçu
  2.  
  3. Vrije Universiteit Amsterdam (VU)
  4.  
  5. Advanced Statistics

Resume

VU Master Health Sciences: Advanced Statistics - Summary lectures and literature: How to analyse clustered and longitudinal data and how to interpret the results?

 10 fois vendu
  • Cours
  • Advanced Statistics
  • Établissement
  • Vrije Universiteit Amsterdam (VU)
  • Book
  • Applied Mixed Model Analysis

All the practical information on clustered data and longitudinal data in a very compact version. How to apply the appropriate analysis step by step? How to interpret the results? - General information about Standard regression analysis (linear and logistic), Clustered data and Longitudinal data. Ba...

[Montrer plus]

Aperçu 6 sur 30  pages

Infos sur le Document

  • Non
  • 1 t/m 9
  • 3 décembre 2021
  • 30
  • 2021/2022
  • Resume

Sujets

  • mixed model analysis
  • longitudinal data
  • gee analysis
  • mma
  • gee
  • growth curve analysis
  • latent
  • clustered data
  • spss
  • stata
  • advanced statistics
  • statistics
  • logistic
  • linear
  • longitudinal
  • multilevel
  • 2 log likelih

Livre connecté

book image

Titre de l’ouvrage:

Auteur(s):

  • Édition:
  • ISBN:
  • Édition:

École, étude et sujet

  • Vrije Universiteit Amsterdam (VU)
  • Health Sciences Master
  • Advanced Statistics
Tous les documents sur ce sujet (1)

Vendeur

avatar-seller
Student1064

Avis reçus

Aperçu du contenu

Standard regression analysis (SRA)
Linear regression analysis
- Describes relationship by fitting line to observed data.
- Uses straight line (logistic/non-linear models use a curved line).
- Estimating how dependent variable changes as independent variable(s) change.


- Y = predicted value dependent for any given value of the independent variable
- B0 = intercept
Predicted value Y when X = 0
- B1 = regression coefficient
How much Y changes as X increases
Denotes magnitude of change in Y
- X = independent variable
- E = error of the estimate
The amount of variation in the estimate of the regression coefficient

OBTAINING REGRESSION LINE
- Least square means = finding best fit for data set points by minimizing the sum of residuals of
points from the plotted curve.

3 steps:
1. Square each residual
2. Sum all squared residuals
3. Minimize the total of the squared values

ASSUMPTIONS
1. Linear relationship between X and Y
2. Normal distribution residuals
3. Homoscedasticity residuals
4. Independent observations




- Independent-samples T-test = to compare means between 2 unrelated groups on the same
continuous Y.
- One-way ANOVA = to compare means between >3 unrelated groups on the same continuous Y.

LINEAR REGRESSION ANALYSIS OUTCOME

,Logistic regression analysis
- Describes relationship between binary Y and >1 other covariates (X).
- To fit a line between observations Y is transformed into the logarithm of the odds (ln(odds)).
- E-power of b1 = odds ratio (Exp(B)).


- Ln(odds of …) = the natural log of the odds of the outcome
- B0 = intercept
The natural log of the odds of the outcome when X = 0
- B1 = regression coefficient
How much ln(odds of …) changes when X changes with 1 unit
 taking the E-power of b1 gives the odds ratio (more easy to interpret)

LOGISTIC REGRESSION ANALYSIS OUTCOME




Confounding
- Confounding = a distortion that modifies an relationship between exposure and outcome, because
the factor is associated with both exposure and outcome.
- Re-assess b1 after adding the potential confounder (b2) into the model.
 compare crude b1 and adjusted b1 and calculate the percentage difference in the b1

Crude b1 Adjusted b1
Logistic -0.461 -0.388
Linear 2.149 2.212

Calculation logistic
-0.% = -0.00461
1% = -0.00461
-0.388 / -0.00461 = 84
100 – 84 = 16%  >10% there is confounding by
sex

Calculation linear
2.% = 0.02149
1% = 0.02149
2..02149 = 102.9
100 – 102.9 = -2.9%  <10% there is no confounding by sex

,Effect modification
- Effect modification = when magnitude of effect exposure (X) on Y differs between the level of the
third variable.
 exposure having different effects
- Assess p-value of interaction term (third variable).
- If significant, results of effect should be reported separately for the different subgroups.

,Basic principles linear mixed model
analysis
- Linear MMA extended version of linear regression analysis.
- Clustering is present when a set of objects group in such a way that objects in the same group (=
cluster) are more similar to each other than to those in other clusters.
- Observations within clusters are correlated with each other.
- You have to take this into account in your analysis with MMA.




Similar approach in regression models with clustering:
1. Intercept (u)
2. Slope (uk)
3. Intercept and slope (ukj)

General idea MMA – 3 steps
1. Estimate intercepts(/slopes) for different groups
2. Draw a normal distribution over the different intercepts(/slopes)
3. Estimate the variance of that normal distribution

Covariance between random slope & random
intercept
- Also known as the covariance (interaction) between the random slope and
random intercept.

1. Negative covariance
Indicates inverse relationship
For levels with relatively high intercept, a relatively low slope is observed
2. Positive covariance
Indicates same relationship
4. For levels with relatively high intercept, a relatively high slope is observed

Intraclass Correlation Coefficient (ICC) – ICC as
indicator
- ICC = indication average correlation of observations of subjects living in the same cluster.
- Indicates how strong units in the same group resemble each other (correlation).
- When calculating ICC with a model that includes an X variable, remaining variance is lower.
- Pure ICC calculated with intercept-only model (model without X).

,Variance used as explanation (specific application MMA)
- Using random effects for explanation differences.
- % of the difference in Y between the levels of the cluster is explained by X.
- Calculate with random intercept of the intercept-only model and the random intercept of the
model with X.

, Example linear MM
Explained with cross-sectional cohort study investigating the relationship between X (physical activity = PA)
and Y (health).
- Two-level structure
Subject = lowest level
Neighbourhood (NBH) = highest level
- Linear regression analysis should adjust for NBH by MMA.

1. Naïve linear MMA
- Without an adjustment for NBH




General information
- MMA without adjustment
- Log likelihood
- Number of observations
Fixed part
- Activity = b1, with standard error (S.E.), z-value, corresponding p-value and 95% CI estimated
around the b1.
Difference in health when there is 1 unit difference in PA
- _cons = intercept
Value of health when PA equals 0
Random part
- Var(Residual) = residual variance (the error variance/unexplained variance).
Because it’s a naïve model, random part only contains variance of the residual.

2. Add random intercept to the model
- Adding a random intercept on cluster level to the model.
- To adjust for NBH level.

Les avantages d'acheter des résumés chez Stuvia:

Qualité garantie par les avis des clients

Qualité garantie par les avis des clients

Les clients de Stuvia ont évalués plus de 700 000 résumés. C'est comme ça que vous savez que vous achetez les meilleurs documents.

L’achat facile et rapide

L’achat facile et rapide

Vous pouvez payer rapidement avec iDeal, carte de crédit ou Stuvia-crédit pour les résumés. Il n'y a pas d'adhésion nécessaire.

Focus sur l’essentiel

Focus sur l’essentiel

Vos camarades écrivent eux-mêmes les notes d’étude, c’est pourquoi les documents sont toujours fiables et à jour. Cela garantit que vous arrivez rapidement au coeur du matériel.

Foire aux questions

Qu'est-ce que j'obtiens en achetant ce document ?

Vous obtenez un PDF, disponible immédiatement après votre achat. Le document acheté est accessible à tout moment, n'importe où et indéfiniment via votre profil.

Garantie de remboursement : comment ça marche ?

Notre garantie de satisfaction garantit que vous trouverez toujours un document d'étude qui vous convient. Vous remplissez un formulaire et notre équipe du service client s'occupe du reste.

Auprès de qui est-ce que j'achète ce résumé ?

Stuvia est une place de marché. Alors, vous n'achetez donc pas ce document chez nous, mais auprès du vendeur Student1064. Stuvia facilite les paiements au vendeur.

Est-ce que j'aurai un abonnement?

Non, vous n'achetez ce résumé que pour 7,39 €. Vous n'êtes lié à rien après votre achat.

Peut-on faire confiance à Stuvia ?

4.6 étoiles sur Google & Trustpilot (+1000 avis)

66184 résumés ont été vendus ces 30 derniers jours

Fondée en 2010, la référence pour acheter des résumés depuis déjà 15 ans

Commencez à vendre!
7,39 €  10x  vendu
  • (0)
Ajouter au panier
Ajouté