Questions tagged [variance-decomposition]
A decomposition of variance explained by a model into additive contributions from each predictor.
45 questions
0 votes
0 answers
47 views
FEVD does not sum up to 100
I am conducting a time series analysis using a VAR model in Python (statsmodels) to study the interactions between macro-financial indicators in Vietnam. I am using daily data and my dataset includes ...
4 votes
1 answer
497 views
Law of total variance with two conditioning variables
The law of total variance, in its standard form, states that $$ \mathrm{Var}\,[Y] = \mathrm E\, [\mathrm{Var}\,[Y \mid X]] + \mathrm{Var}\, [\mathrm E\,[Y \mid X]]. \tag{1}\label{1} $$ Instead of one ...
- 1,285
2 votes
1 answer
139 views
How to compare the contribution between a fixed effect and a random effect in a multilevel model?
Suppose I have a data set with three variables: $y$, an outcome I want to understand, $x$, a variable which covaries with $x$ but we can assume to be measured perfectly, and $g$, a grouping variable ...
- 1,087
4 votes
3 answers
480 views
Multiple linear regression: Partial effects interpretation of parameters at/near collinearity
I am looking at the simplest example of multiple linear regression model: \begin{equation}\label{linreg} Y = \beta_1 X_1 + \beta_2 X_2 + \varepsilon, \end{equation} I am interested in the ...
- 764
3 votes
1 answer
221 views
Why does variance decomposition not exist in logistic regression?
I'm quite confused when reading some materials, it says the reason we use pseudo R-square in logistic regression is that the variance decomposition cannot hold. I get the idea that the variance of $y$ ...
- 31
0 votes
1 answer
154 views
What is the variance decomposition method?
For $i = 1, \ldots, m$ and $j = 1, \ldots , n$ we have observations $x_{ij}$. We can assume that $$ x_{ij} = y_{i} + z_{ij}, \qquad y_{i} \sim \mathcal{N}(\mu_{y},\sigma_{y}^{2}), \quad z_{ij} \sim \...
- 101
1 vote
1 answer
86 views
How can a Bayesian linear hierarchical random intercept model with normally distributed priors for coefficients represent a non-normally DV?
Suppose you have a hierarchical random intercept model with a dependent variable that is zero inflated. The link function is linear and the priors for the coefficients are normally distributed. In ...
3 votes
0 answers
845 views
Blinder-Oaxaca decomposition, logistic regression and unbalanced dataset: fitted probabilities numerically 0 or 1 occurred
I have a binary y outcome, a dummy variable gender for gender, and a set of covariates x (...
- 3,403
2 votes
1 answer
320 views
How to decompose the determinants of an outcome
I was looking at this graph about the gender pay gap. I have a similar problem. I have an binary outcome $Y$, a binary covariate $D$ for gender (with 1 meaning female), and a series of covariates $X_1$...
- 3,403
1 vote
0 answers
273 views
Within- and between-group variance does not add up to the total variance [duplicate]
The variance is additively decomposable (Shorrocks 1982). That is, when decomposing it into within- and between-group components, the within and between-group component should add up to the total ...
- 215
2 votes
0 answers
176 views
Variance decomposition for time series of weighted averages
I face the following problem: I have a time series $x_t,\;t=1,...,T$ where each value $x_t$ is a weighted average for various groups, that is $x_t = \sum_{i=1}^Nf_{it}*x_{it}$ with $\sum_{i=1}^Nf_{it}=...
- 106
0 votes
1 answer
446 views
Using a Panel Regression to Explain Aggregate Effects/Interpreting the "size" of panel regression coefficients
Suppose I've estimated a two way fixed effects regression using unit level panel data: $$ y_{it} = \beta_1x_{it} +\beta_2x_{it}H_{it} + \beta_3H_{it} + a_i + \gamma_{t}+\epsilon_{it} $$ with $(a_i, \...
- 3
1 vote
1 answer
2k views
Decomposing the coefficient of variation into within- and between- group components
I know that the variance can be decomposed into within- and between-group components. Let $Y$ be a vector and $Y_{i}$ and be a subgroup of $Y$ and we have $I$ subgroups, then $Var_{T}(Y)=Var_{W}(Y)+...
- 215
0 votes
1 answer
559 views
Calculate change in odds and Pseudo-R-Squared with R [duplicate]
I have a question that touches on both technical solutions in R and statistics. I have a huge dataset with 2,400 respondents in total. I performed a logistic regression in order to analyze views on ...
- 101
2 votes
0 answers
1k views
Linear Mixed Regression Variance Decomposition
I'm pretty new to R and was hoping to get some advice on variance decomposition in mixed linear models. Similar to this question; How to estimate variance components with lmer for models with random ...
- 23