All Questions
31,016 questions
0
votes
0
answers
7
views
Accidentally including the same variable twice in a regression model?
I have this mixed effects GAM regression in R (mgcv):
...
- 667
0
votes
0
answers
8
views
How to Test and Implement Industry and Year Fixed Effects vs. POLS in Panel Data Analysis?
I am new to econometric modelling.
I am working with panel data spanning 8 years, including 155 units (REITs) distributed across 8 different industries. I aim to apply industry fixed effects and year ...
- 1
2
votes
0
answers
154
views
Deriving the normal equations by differentiating the matrix form
So, if we have $RSS(\beta) = (\mathbf{y}-\mathbf{X}\beta)^T(\mathbf{y}-\mathbf{X}\beta)$ and differentiatiate with respect to $\beta$, and set it to zero in order to minimize it, how do we get $\...
0
votes
0
answers
52
views
Have I correctly derived the iterative updates for weighted least squares?
I have an exercise where I have to derive the both $w_i^{(m-1)}$ and $z_i^{(m-1)}$ from the iterative weighted least squared updating equation $b^{(m)} = \left( X^\top W^{(m-1)} X \right)^{-1} X^\top ...
- 101
0
votes
1
answer
21
views
What are the differences between PS-match and adjusted Cox regression?
This is more like an extension of the following question: Propensity Score Matching with Cox Regression
I am wondering what are the differences between these:
matching patients with PS and running ...
- 581
0
votes
1
answer
37
views
Using a mixed effects model to understand the population?
I have this repeated measures data in R (simulated) that represents multiple regions in the same country (i.e. multiple measurements over time for each region - assume regions are measured at the same ...
- 667
0
votes
0
answers
24
views
Does incorporating lower bound on dependent variable improve estimation?
Suppose we flip a coin $n$ times, where $Z_k = 1$ if heads otherwise 0, and $Y = \sum_{k=1}^{n} Z_k$ is the total number of heads. There are two types of coins: silver and gold, which we can observe ...
- 498
0
votes
1
answer
20
views
Interaction term negative when both its components are positive?
I examined the effect of labour cost, labour quality and their interaction (cost*quality) on FDI, but I got positive coff. of both components and negative coff. of interaction term. How could to ...
- 3
0
votes
0
answers
13
views
Modelling the change between pairs of points vs longitudinal regression?
This is the basic Longitudinal Model (Mixed Effects), used to account for correlations between repeated measures in groups of individuals:
$$ y_{it} = X_{it}\beta + Z_{it}b_i + \epsilon_{it} $$
Where:...
- 433
2
votes
1
answer
35
views
Non-Parametric Regression: Seminal Papers
This question is mostly a question for references and is not objectively answerable as recommendations for literature inherently depend on personal preference. I hope it is still ok.
I would like to ...
0
votes
0
answers
8
views
In heckman model can i use dependent variable in selection equation as independent variable in outcome equation? [closed]
I have included above said variable as dependent in selection and independent in outcome equation.. But in results the choosen and most importent variable got omitted.. It is showing ommitted because ...
- 1
1
vote
0
answers
53
views
Why do machine learning courses on regression mostly focus on gradient descient although we have the closed form estimator $(X'X)^{-1}X'Y$? [duplicate]
In many online machine learning courses and videos(such as Andrew Ng's coursera course), when it comes to regression (for example regressing $Y$ on features $X$), althouth we have the closed form ...
- 3,050
0
votes
0
answers
5
views
Regression table for fixest and plm models?
I am working with fixest and plm models. I usally use fixest's etable for fixest objects and stargazer for plm. The problem is ...
- 183
1
vote
1
answer
30
views
Identifying Poorly Forecastable Time Series Using tsfeatures
I am working on a problem involving the identification of poorly forecastable time series using features extracted with the tsfeatures library by Rob J. Hyndman. Below are the key details about my ...
1
vote
0
answers
26
views
Linear Regression and ridge regression resulting in similar coefficients
The coefficients I am obtaining from linear regression are similar to the ridge regression. I have tried to understand where I am getting wrong I could not figure out. It would be very helpful if ...
2
votes
0
answers
23
views
Taking partial derivatives of regression models
For a general statistical model with function $f(x_1, x_2, x_3)$, the partial effect of $x_1$ is what we get when we take the partial derivative while holding $x_2$ and $x_3$ fixed at specific values (...
- 667
0
votes
0
answers
17
views
Can I use a series of linear regressions to perform post-hoc tests on survey data in R following a comparison of means for 3+ variables?
Thomas Lumley explained within this StackOverflow post that the survey package's regTermTest() function can be used to compare ...
- 101
5
votes
1
answer
115
views
+50
Is there a standard way to calculate marginal effects?
My question is about how to set the values of the other predictor variables when calculating the marginal effects for a regression model.
I have a GAM regression model (response is continuous between ...
- 667
1
vote
0
answers
20
views
Regression models that depend on outputs of other regression models
There is a milk factory with the following variables available at the weekly level (i.e. data at the end of each week) and orders are finished on a first in first out basis:
total incoming orders (...
- 141
1
vote
1
answer
49
views
GAM Regression: Interactions vs Main Effects?
I have a GAM regression model (response is between 0 and 1):
$$ g(\mathbb{E}[Y_i]) = \beta_0 + f_1(t_i, x1_i) + f_2(t_i, x2_i) $$
$g(\cdot)$ is the logit link function $g(p) = \log(\frac{p}{1-p})$
$...
- 667
0
votes
0
answers
11
views
What does the subscript "F" in the glmnet multi-task regression optimization problem define?
I was planning on using the glmnet implementation of ElasticNet multi-target regression. I was examining the documentation to accurately describe the technique in a manuscriptI am writing.
The ...
- 21
6
votes
1
answer
272
views
Does it mean that we don't need a normal assumption for using sandwich estimator in normal linear regression?
According to this post,
the blogger uses the theory of estimating equations to construct the robust sandwich variance estimator.
In this post, it said that:
Now we ...
- 364
3
votes
2
answers
154
views
Will marginal effects for a logit link also be between 0-1?
I have a question on marginal effects.
I have some data in R (response is between 0 and 100, t is a time variable, predictors are continuous and greater than 0):
...
- 667
0
votes
0
answers
41
views
Quantile regression with a pre-existing cutoff point in the raw score of dependent variable
This is probably a dumb question, but I have never used quantile regression so I'm unsure.
I have the following situation: there is a continuous, numerical self-report measure with a certain, pre-...
- 2,825
0
votes
0
answers
18
views
How to plot hazard ratio from coxph model with tt() term [closed]
This question is related to another that I posted here:
How to visually assess tt() suitability in coxph
If we have a time-varying HR that arises from a time-dependent coefficient because we have ...
- 857
0
votes
0
answers
31
views
How to visually assess tt() suitability in coxph? [closed]
For this example, I am taking cues from the time-dependent survival vignette.
I am interested in understanding how to assess the suitability of a covariate-time interaction using ...
- 857
0
votes
0
answers
16
views
Negative Lambda in box/cox: Caveats and Interpretation [closed]
I have a highly-skewed DV which has not responded well to standard transformations (square root, cubed root, log or ln, hyperbolic arcsine).I do get some reduction in skew, but not enough to normalize ...
2
votes
1
answer
93
views
+50
Interpretation of estimated effect of a predictor variable when it does not vary within a sampling unit
Let us consider the following scenario. John is soon going to quiz his students and he is interested to know if the font type with which a quiz is formatted is associated with the performance of the ...
0
votes
1
answer
41
views
Does H0:beta_1=0 in SLR assume normality of outcome which MIGHT NOT BE TRUE AT ALL?
Theoretical Simple Linear Regression Model under the Null Hypothesis
The theoretical simple linear regression model under the Null Hypothesis ($\beta_1 = 0$) is:
$$
Y_i = \beta_0 + \epsilon
$$
where $\...
- 1
0
votes
0
answers
29
views
identification strategy soccer match
I have data on what minute a player was given a yellow card in a soccer match. I also know how long a particular player was on the pitch (i.e. if he was on the pitch from the start or substituted). I ...
- 155
-2
votes
0
answers
21
views
Filtering on Fortune 1000 companies [closed]
I'm looking for stock price data for Fortune 1000 firms. Does anyone know whether there is an easy way to filter specifically for Fortune 1000 companies in Compustat (or another database)? I.e., is ...
- 1
0
votes
0
answers
39
views
How to identify the best use of the logit link function in regression?
I am confused about how to identify the best use of the logit link function in regression problems.
Here is my current understanding of the topic. My confusion stems from the fact that for continuous ...
- 141
1
vote
1
answer
49
views
IV Rank/Relevance Condition Linear Algebra Intuition
Consider the following econometric model (IV) : $Y_1 = X'\beta + e$, where $Y_1 \in \mathbb{R}$ is some outcome variable of interest, and we have a set of regressors $X = \begin{bmatrix} Z_1 \\ Y_2 \...
- 101
7
votes
2
answers
149
views
+50
Standardizing variables for a regression model vs weights in a regression model?
I have a longitudinal GAM (General Additive Model) regression in R.
Here is the general form of the model and data (response is between 0 and 1). All variables are calculated at the state level (e.g. ...
- 667
0
votes
1
answer
15
views
plm and time effects
To estimate a panel data using the within model, we use the plm package
For example if I have two variables, y, x we have
...
newfinder
-1
votes
0
answers
27
views
Every regression problem, can be cast into classification problem. How about the vice versa? [closed]
If regression is about estimating the continuous value of ground truth and predicted values, and continuous values can be cast into discrete values, then discrete values can be labeled as ...
0
votes
0
answers
29
views
Why bother with k-means number of clusters? Why not generate them all and see which one works? [closed]
I'm a sociologist with a CS background. I'm analyzing longitudinal data and I'm not up to speed with the statistical lingo around the whole thing. I'm trying to figure out the statistical names of the ...
0
votes
0
answers
36
views
Confusion over vector notation in regression paper
I'm confused about the notation I've seen occasionally in some econ papers. Here is an example,
$Y_i=\alpha_0+\alpha_1 G_i+\alpha_2 G_i\times H_i+\alpha_3 A_i+\alpha_4 M_i+ \alpha_5 H_i+\gamma_{s(i)}+\...
- 187
1
vote
1
answer
33
views
Modeling longitidunal relationships without estimating indirect effects
I'm analyzing a three-time point study using structural equation modeling (SEM). I only care about the direct effects of my variables. Is it okay to only model those direct paths, or should I include ...
- 13
0
votes
0
answers
16
views
Computation question for ANOVA with matrices
I'm currently studying for finals and have come across this question that I'm having trouble approaching. The linear regression model is defined as always (in this particular scenario with 2 ...
- 181
2
votes
1
answer
27
views
Prediction from regression equation in opposite direction [duplicate]
Let say I have following simple linear regression
$y = \alpha + \beta X + \epsilon$
Here $X$ is ordered categorical variable, let say with categories represented by 1 to 12.
Now in typical regression ...
- 415
1
vote
2
answers
100
views
Data passing all linearity assumptions except normality. What should my next steps be?
I am required to build models for a series of datasets which all have the same issue.
Edit: An example of one of my datasets is displayed in the image below, with a linear regression line fitted to it....
- 29
0
votes
0
answers
6
views
Analysis of GBS before and after COVID-19 while adjusting for confounding variables
I am new to statistics and would appreciate the help! I am using SPSS and am working on a project where I want to analyze the impact of COVID-19 on Group B Streptococcus (GBS). I therefore have 4 ...
- 1
1
vote
1
answer
48
views
Estimating mixed model with identical response value but different covariate values within a pair
Say we have a dataset with individuals. Each individual performed a task, either in solo or with another individual (variable condition), and we measured the ...
- 65
1
vote
0
answers
11
views
Interpreting an exponentiated coefficient from a regression of a log-transformed outcome when coefficient is < 1 [duplicate]
Plenty of stuff online (see here and here) about what to do when you log transform an outcome (exponentiate regression coefficients) and how to interpret the coefficients (1-coefficient*100 = ...
- 2,187
1
vote
0
answers
34
views
Least Absolute Deviations – Geometric Intuition
I've recently been exposed to the geometric intuition regarding Least Squares (OLS) regression:
The vector of the outcome variable $Y$, is not not in the linear span of $X_1, X_2, ..., X_{p-1}$:
The ...
- 31
2
votes
1
answer
30
views
Alternative to R-squared when calculating glm model using survey package
I'm writing my master thesis using the European Social Survey. The data requires weighting, so therefore I have to use the survey package and its ...
- 21
0
votes
0
answers
31
views
In the R function rank_trace, does there exist a common threshold value to accept a rank?
In the R package rrr (document) which is used for Reduced Rank Regression, there is a function ...
- 113
0
votes
1
answer
29
views
Modelling presence/absence data with PC scores
I have some data of animal counts vs. PCA scores of environmental variables that looks as below. The .5s in the count data are because two people counted and we took an average.
...
- 123
1
vote
1
answer
29
views
INterpretting residuals from a GAMM model
I am trying to run GAMs using binomial data (link=logit) in r with the mgcv package. This is to attempt to describe which variables describe species presence using presence (1) and absence (0) data as ...
- 41