Questions tagged [multilevel-analysis]

Statistical analysis of datasets comprising several levels of hierarchy (e.g., students nested in classes nested in schools or hierarchical forecasting). For questions about mixed models use [mixed-model] tag. For nested random effects, use [nested-data].

Filter by
Sorted by
Tagged with
1 vote
0 answers
5 views

Structural or sensitivity analysis of multivariate time series with multiple subjects

Sorry if this isn't explained in the best way. I have very basic knowledge of time series analysis so my question may sound very simplistic or might be missing the big picture of this type of analysis....
user avatar
0 votes
0 answers
11 views

Calculating sampling variance in multilevel meta-analysis

I'd like to conduct a three-level meta-analysis using R. I see that several papers recommend calculating the level 1 sampling variance using Cheung's 2014 formula 14. However, I'm wondering if it is ...
user avatar
  • 1
0 votes
0 answers
13 views

When will a cross-level interaction increase the slope variance?

Heck et al (2012) give the following model $Y_{ij} = \gamma_{00} + \gamma_{01}\text{ses_mean}_j + \gamma_{02}\text{pro4yrc}_j + \gamma_{03}\text{public}_j + \gamma_{10}\text{ses}_{ij} + u_{0j} + \...
1 vote
0 answers
6 views

Do I need multilevel modeling for the data with a small number of clusters + an imbalance in cluster sizes?

My data is organized in nested levels: 48 children (individual-level) nested in 3 schools(cluster-level): n=30 in cluster 1, n=15 in cluster 2, and n=3 in cluster 3. I simply wonder if I will need to ...
user avatar
  • 11
0 votes
0 answers
10 views

What does it mean if a predictor was significant in a non-mixed-effects model, but no longer significant after using a mixed-effects model?

Say you are working with college education data and are aiming to predict students' pass or fail of a specific course. So you first try a logistic regression model that includes 5 predictors that are ...
user avatar
  • 41
0 votes
0 answers
12 views

How to run a multilevel regression model with crossed random effects: items and participants?

I am trying to run a multilevel regression for my study: I have two random effects; participants (97) and items (which are the 20 words used in the study) Each participant had to spell the same words. ...
user avatar
2 votes
0 answers
17 views

is ok to use the same variable in two hierarchies?

I have a dataset with two levels: the city level and person level, at the city level i have the categorical variable "poverty level", in my research estimating the effect of the "...
user avatar
0 votes
0 answers
8 views

Setting up a stacked dataframe for multilevel model analysis in R

I need to set up my data frame into a stacked or long form in order to do a random crossed effects multilevel analysis. My random factors are participants and items (words in the assessment) My fixed ...
user avatar
0 votes
0 answers
6 views

Multilevel models significance

For example, in the OLS regression (not multilevel), we have R^2 and p-value for F test. This p value indicates whether R^2 is significance or not. In multilevel models, we have R^2 as well (marginal/...
user avatar
  • 109
0 votes
0 answers
8 views

Multivariate multilevel ordinal regression

I am looking to estimate a multivariate multilevel ordinal regression model (preferably via the ordered logit) in $R$ using non-bayesian methods. Is anyone aware of a package that does this? If not, I ...
user avatar
0 votes
0 answers
10 views

Sample size for interaction term in HLM

I am running a hierarchical ordinal regression model with 41 subjects evaluated at 5-time points each. Besides a single level-1 time-varying covariate, I only have two other explanatory variables - ...
user avatar
1 vote
0 answers
18 views

glm() vs. glmer() on repeated measures

Say I am curious about whether the relationship between class rank (rank) and passing a final test (pass) is dependent upon days until summer vacation (days until summer). In my dataset, I have Mary, ...
user avatar
  • 131
0 votes
0 answers
22 views

Multiplying (or averaging) effect of independent Bayes Factors

I want to know how to combine the effect of Bayes Factors calculated on subsets of a dataset. Note, this is not the case of replication BF, where I have, say a BF from a previous study (which acts ...
user avatar
  • 151
0 votes
0 answers
53 views

Can adding a Level 2 predictor impact the Level 1 variance?

Heck et al (2013) p.137 write: [O]ne approach often used is to examine the change in residual variance that occurs by adding predictors within a sequence of models. The analyst begins with the ...
user avatar
0 votes
0 answers
10 views

F statistics in Multilevel modeling

In OLS regression (not multilevel regression) R^2 indicates how much variance in DV can be explained by this model. And, the P value of F-statistics indicates whether this regression equation is ...
user avatar
  • 109
0 votes
0 answers
9 views

Centering in 2 level Multilevel modeling

When two level-1 variables interact with each other, some articles recommend using cluster mean centering, rather than the grand mean centering. However, I do not understand why I should not use grand ...
user avatar
  • 109
1 vote
0 answers
18 views

R: Fable - forecasting hierarchical time series with transformation

I have a hierarchical time series, with two sub-series that have significantly different behaviours. One subseries would definitely benefit from Box-Cox transformation to stabilise the variance. Is ...
user avatar
0 votes
0 answers
9 views

centering issue in the multilevel modeling

I learned that categorical variable should be centered just like continuous variable. I am analyzing two level model, (MLM). case 1. when the categorical variable is level 1 predictor, and have ...
user avatar
  • 109
0 votes
0 answers
7 views

Level 3 variance increase when adding a contextual predictor?

I am conducting a multilevel regression analysis on a survey dataset that has three levels: Individuals - Country-Years- Country. I am regressing a series of time varying and time invariant ...
user avatar
2 votes
1 answer
64 views

Post hoc power analysis for multilevel regression analysis

I have a multilevel model with 2 levels (L1 = individuals, at least 710 per country; L2 = countries, 17 total) ...
user avatar
  • 1,448
0 votes
0 answers
11 views

Multilevel modeling, null model, R problem: very large eigenvalues

I found a problem when running the null model of my multilevel modeling in Rstudio. This is the error message that I receive ...
user avatar
2 votes
1 answer
32 views

Can I improve linear model coefficient estimates using group information without working it into model?

I am fitting a linear model in order to predict future observations. The training data consists of about 1000 observations. Each observation comes from one of 10 individuals, which means I have about ...
user avatar
0 votes
0 answers
23 views

Using and Reporting ICC for Aggregated Data About Groups of People

I've designed a survey that consists of multiple questionnaires with Likert-scale questions, each of which targets a specific variable in my research model. The participants of this survey are ...
user avatar
0 votes
1 answer
23 views

Correcting for multiple comparisons in interaction

I have 13 separate hierarchical linear regression models- all models have an interaction term (group by variable of interest) entered as the last step. Since I have 3 dummy coded groups, I have 2 p-...
user avatar
0 votes
0 answers
35 views

Is it possible to implement ANOVA analysis with categorical dependant variable?

I have some survey data which tracks the employment status of individuals over time. I am doing an analysis on the effects of different independant variables on occupation, and one thing I wish to ...
user avatar
  • 101
6 votes
1 answer
135 views

Expected Fisher information isn't positive definite for truncated normal with heteroskedasticity

This question is about having a non-positive-definite expected Fisher information in a normal model in which observations have different dispersions. Consider this simple normal model: $$Y_i \sim N(\...
user avatar
  • 3,491
1 vote
0 answers
22 views

If I'm conducting a matched study with multiple levels (e.g., students within classrooms), what is the appropriate level to match on?

Let's say that I'm conducting a non-randomized matching evaluating of an educational intervention to determine its impact on student grades. While the students receive the intervention, the teachers ...
user avatar
  • 157
0 votes
0 answers
22 views

Centering in Multilevel/ Mixed Modeling (MLM/HLM) Cross-Level Interaction Small Groups Dichotomous Variable

I am having some trouble with the literature on centering in multilevel models. I cannot seem to find a clear answer for my unique situation. I have a multilevel model with approximately 500 level 1 ...
user avatar
  • 1
0 votes
0 answers
7 views

all variables at one of my level is not significant

I am modelling a logistic regression multilevel with 3 levels. Level 1 is individual, level 2 is household characteristic, and level 3 is regional characteristics. The results of the random effect ...
user avatar
0 votes
0 answers
8 views

REML for multilevel modeling comparison

With the very same data, I want to compare the two multilevel models. (these two models differs in both fixed and random effects) so, I am planning simulation study. I will generate the data set, and ...
user avatar
  • 109
0 votes
1 answer
26 views

Account for different amount of observations per participants without using a regression approach?

I am trying to figure out the statistics of a study where I have 60 participants with varying amounts of observation per participant. I am measuring facial muscle activity after a button press over a ...
user avatar
  • 1
0 votes
0 answers
10 views

Hierachical model causing too much shrinkage?

I use population-wide, 12-year data for examining regional differences in rehabilitation use (zero-inflated, lognormally-distributed outcome variable). I analysed the outcome variable with 2 models: <...
user avatar
  • 1,533
1 vote
0 answers
54 views

Multilevel CFA in R (lavaan); Error: No variance within some clusters

I am running a multilevel CFA in R package lavaan (following the process by Dyer, Hanges, & Hall, 2005). I was able to run an individual-level CFA, and examine both the within group covariance ...
user avatar
  • 11
0 votes
0 answers
12 views

Choosing the appropriate model: mlm | panel FE | TSCS

I am a bit unsure what is the right model for assessing overtime variation with unit fixed effects, contingent on context features. The data (n>10000) at hand consists of features of individual ...
user avatar
1 vote
1 answer
24 views

Repeated measures with nested data in R

Previously, I conducted a model like the following. I have repeated measures (Time factor: pre-post) of depressive mood in two different groups (Group factor: neutral and experimental). Each ...
user avatar
1 vote
0 answers
22 views

QQ plot for multilevel logistic regression

I'm running the multilevel logistic regression, and I hope to create a diagnostic plot for the model.But I found the previous qqnorm() function isn't available for multilevel logistic regression. So I ...
user avatar
1 vote
1 answer
49 views

Making binary prediction with GPBoost (or MERF)

My question is regarding this post from 1.5 years ago: Modelling clustered data using boosted regression trees My label is a binary variable (yes/no). Is it possible to use GPBoost / MERF in order to ...
user avatar
  • 13
0 votes
0 answers
4 views

Design effect of a random intercept model-how variance and effective sample size are calculated

I am reading about the random intercept model for the two-level analysis. The effective sample size is defined as neff=Ntotal/(1+(n-1)*ICC), where 1+(n-1)*ICC is also called the design effect. I ...
user avatar
  • 1
2 votes
1 answer
71 views

Meta Analysis in R (metafor): How to interpret a multi-level mixed effects model with intercept and without intercept?

Dears, I am having trouble making sense of two MLMEs I did using rma.mv(). I am using a factor variable with 4 levels as a moderator. The two outputs (once with ...
user avatar
  • 33
0 votes
0 answers
13 views

Is it possible for a hierarchical model to forecast a sub-region with complete lack of data as "borrowing strength"?

I am looking at sales time-series data (weekly from Jan to Dec 2021) which has a natural hierarchical structure by geography. For example, storeA1, storeA2, .... storeA99 located in Neighborhood A, ...
user avatar
  • 1,017
0 votes
0 answers
15 views

How can i check homoscedasticity for Level-2(between)-residuals in a twolevel model?

i have specified a random intercepts and slopes model with a Level 2-predictor for the intercepts on the between level. I have done the estimation with the lme4 package in R. Now, i want to plot the ...
user avatar
  • 1
1 vote
1 answer
24 views

Missing R Squared and ICC for Multilevel Binary Logistic Regression

I am currently assessing if allocation to a rejection scenario and degree of negative beliefs predicts the odds of engaging in a series of behavioural outcomes (e.g. avoiding the person). My first IV (...
user avatar
  • 11
0 votes
0 answers
5 views

Model choice for group-level outcome over time with individual-level predictors

Main question: What is an appropriate model for this? Similar unanswered question here. So far, I've just calculated group-level means of the predictor at each timepoint and run an aggregated ...
user avatar
0 votes
0 answers
14 views

Advice needed on multi-level nature of my data for latent class analysis #multi-level latent class analysis

I have data with a mix of predictor variables, some of them are person level scores (from prospectively collected data), but most of the variables are based on geocodes (e.g. score based on respondent ...
user avatar
0 votes
0 answers
8 views

How to run and interpret Multilevel Models in SPSS?

I'm totally new to using SPSS and am trying to understand how to run and interpret multilevel models (which I'm also fairly new to)... I have data for 4 groups of around 200 participants each, and all ...
user avatar
1 vote
1 answer
26 views

3-level meta-analysis: Level 3 (study level) has a variance of 0, should I drop this level?

I am doing a 3-level meta-analysis. First, I fit a theoretically-driven model, where level 1 is the sampling variance, level 2 is the effect level, level 3 is the study level. This method is well ...
user avatar
  • 53
0 votes
0 answers
19 views

How to adjust p-value for both multi-level and multi-variate comparison?

I have a naive statistical question: Do I need to adjust p-values two times when I perform a multi-level and multi-variate comparison? Suppose I have a data frame df...
user avatar
1 vote
0 answers
25 views

Three level multilevel model in r

I am conducting an analysis looking at trust across a range of 30 countries between 2002-2018. I am using repeated cross sectional survey data. My data is structured so that individuals(level 1) are ...
user avatar
0 votes
0 answers
17 views

How to interpret linear mixed effects model for non-changing binary fixed effect?

Let's assume we have a model looking at the effect of gender (m/f) on test scores (0-100). Data is being pulled from multiple students across multiple universities, with the restriction that a student ...
user avatar
  • 125
0 votes
0 answers
6 views

Moderation analysis with DV and IV having two measuring points: mixed-effects model?

I have been struggling with the plan for the statistical analyses of my research project for my master's. In essence, it is a moderation analysis. However, both the DV and the IV are measured twice (...
user avatar
  • 1

1
2 3 4 5
34