Count data are non-negative integers representing whole amounts. When such data are the dependent variable in a regression, Poisson or negative binomial regression may be appropriate methods. One common problem is "zero-inflation" (where the proportion of zero values is greater than predicted by a ...

learn more… | top users | synonyms

1
vote
0answers
23 views

Odd looking residual plot - not sure what transform to use if any

I am concerned about the residual plot shown. The (count) data are over-dispersed, with about 40% 0s, median is 2, maximum is 300 or so. I am not sure what how to proceed with this - it is not ...
0
votes
0answers
40 views

Should I ignore negative prediction values?

I have the following time series of count data: ...
0
votes
1answer
27 views

Logistic Growth models for Count Data

I have a dataset of monthly ridership figures by transit route from 2007 to 2015. I am analyzing this data in R. When I go to predict on a new dataset with step increases in trips (ie 1,2,3,etc.) ...
0
votes
0answers
2 views

Modeling binary outcome variable using cumulative exposures as independent variable

I'm stuck on a modeling question. I have to predict a binary outcome variable based on subjects' cumulative exposures to a stimulus over time. Not all subjects are exposed to the stimulus the same ...
0
votes
0answers
13 views

How to use multilevel analysis (MLM) in SPSS when I have 1 DV (frequency of absenteeism) and multiple IVs (more then ten) over three levels?

My aim is to analyze data in SPSS from an employee survey (approx 2000 subjects) and link this data to absenteeism. I think I should use multilevel analysis, but I am not experienced with MLM. DV = ...
3
votes
1answer
110 views

Non-normally distributed data - Box-Cox transformation?

I have data that is not normally distributed. The problem seems to be that there are too many of one value relative to other values. What I have tried to make data normal: I have tried a log ...
2
votes
1answer
52 views

Creating a probability distribution that is truncated skewed

I have a dataset I want to use to generate a probability distribution. The distribution is skewed and can only include positive integers. I've tried normal (both skewed and truncated, although I ...
1
vote
0answers
20 views

comparing multiple proportions over time

I have a dataset of 22 fish species in a fish market, sampled monthly, from 2007 to 2011. I want to see if there is a statistically significant change in their relative proportions over time. I first ...
1
vote
1answer
43 views

PCA on count-based data

I'm looking to do a PCA analysis on count based data itself rather than averages. I'm hoping this will help for variable observation depths; for example, 3/4 reads is not really equivalent to 15/20. ...
0
votes
0answers
11 views

Comparing hurdle models to negative binomial models

I'm trying to compare the AIC or log-likelihood of a negative binomial GLM to a hurdle type approach, consisting of a binomial GLM for the presence/absence of a count and the counts modelled with a ...
2
votes
0answers
29 views

Modelling overdispersed counts - past negative binomial

I'm modelling overdispersed counts. I began using a GLM with Poisson error structure, then moved to quasi-Poisson, and then finally negative binomial. The residuals versus fitted values plot is still ...
2
votes
0answers
36 views

Estimate lambda for panel count data

I have panel count data for $F$ firms across $I$ years, so observe counts $C_{f,i}$ for $f \in \{1,...,F\}$, $i \in \{1,...,I\}$. I want to model the data as a poisson process. With increasing ...
0
votes
1answer
28 views

Normalize counts over time without adding excess variability

I'm not quite sure if this is possible to do, but I have count data that I would like to put on a similar scale in order to compare it. We are tracking where people go, and want to see if people who ...
0
votes
1answer
34 views

What regression model or test is better for modeling commuting data?

I have count of people commuting from an origin to a destination as my response/dependent variable. The independent variables are travel time (in minutes) and travel distance (in kilometers). I have ...
1
vote
0answers
21 views

Correlation between counts

I'm trying on attendance data to find relationships between where people go. For example, say I have information on users A, B, C & D and what kinds of places they've visited in the past few ...
1
vote
0answers
21 views

Software or workarounds for vector autoregression on count data?

I have a research question that fits nicely in a VAR framework but my data are count data. They are much closer to a Poisson or negative binomial distribution than they are to a normal. This will ...
0
votes
0answers
48 views

What significance test to use for count data?

I have TRANSFAC data of number of transcrption factor binding sites(TFBS) for each gene. I have ~1100 Transcription factors (TFs) and 2 sets of genes: 18 genes belonging to Pigmentation AND 5 genes ...
0
votes
0answers
17 views

Logarithmic offset + Scaling Counts + Multiplying by Scalar + Poisson Regression

I am working with a contingency table where I know that the population sizes from which the counts are derived are very different. From what I have been able to gather, an appropriate way to adjust ...
1
vote
0answers
32 views

Should predictions with negative binomial regression only produce integers?

I have a dataset consisting of about 600 observations. Each observation has around 100 attributes. One of the attributes I want to predict. Since the attribute that I want to predict can only have ...
1
vote
0answers
65 views

Regression models to only predict integers (instead of floating point numbers)?

I have a dataset that consists of about 50 different attributes. One of these attributes I want to predict, using the other attributes as features. The values of the attribute that I want to predict ...
1
vote
1answer
32 views

Alternatives to Chi-Squared for Single Categorical Outcome and Single Categorical Predictor w/counts for factors [R]

I am from an applied background, where X2 and G-tests are the default ways to analyze count data (default as in, until today, I had no idea there were other ways, as I was only taught these methods). ...
3
votes
1answer
112 views

Statistical comparison between binomial distributions, two groups and many trials

I have a data set with two groups. I have multiple trials with different Ns in each trial, and different numbers of trials for each group. In each trial, for each ...
1
vote
1answer
95 views

Using linear regression for count data - will this introduce bias?

Say I am fitting a model to Poisson count data, but I am only interested in estimating the mean of the count variable. I understand a ordinary linear regression is a good approximation when the ...
0
votes
0answers
18 views

Percentage/Mixture data with many zeros

My dependent variable is continuous. My independent variables can be looked at in two ways. In the first, they are a bunch of count data with a large cluster at zero. In the second way, we can ...
1
vote
0answers
33 views

Factor Analysis of Count Data

I am new to factor analysis. I inherited a project at work from another team. They took 9 variables that are all Poisson-distributed count random variables and ran a "regular" factor analysis in ...
3
votes
1answer
67 views

Use of Poisson distribution to analyse distribution of individuals in space

Dytham 2010 suggests using the Poisson distribution to establish whether individuals are evenly distributed in space. Say we end up with a map of individuals in a study site that looks like the ...
2
votes
2answers
118 views

Am I breaking the assumptions of the Poisson distribution?

The Poisson distribution arises when events are counted within a specified interval. I've recorded the number of events each month (I'll not discuss what these events represent). This appears to meet ...
2
votes
2answers
62 views

Help on deciding how to perform a regression analysis on my data, and why? SPSS, Count model, Panel data

I am currently in the process of analyzing some data for my master thesis, however I have had little statistical teaching, and none regarding panel data regressions. My dependent variable: Count data ...
0
votes
0answers
12 views

Is a gaussian mixture appropriate for TF-IDF?

I'm trying to fit multivariate mixtures of TF-IDF scaled variables. So these variables are weighted count proportions of words in a corpus (relative frequency in document weighted by the log of ...
10
votes
1answer
606 views

When to use Poisson v. Geometric v. Negative Binomial; GLM, the Three Count Variable Distributions

I'm trying to layout for myself when it's appropriate to use which regression type (geometric, poisson, negative binomial) with count data, within the GLM (only 3 of the 8 GLM distributions are used ...
0
votes
0answers
23 views

Modeling count data [duplicate]

Why do count data need to be modeled differently from the standard linear regression model ? Any simple example will suffice.
11
votes
1answer
325 views

significance of difference between two counts

Is there a way to determine whether a difference between a count of road accidents at time 1 is significantly different from a count at time 2? I have found different methods for determining the ...
0
votes
2answers
35 views

how best test for differences among two samples of abundance (count) data measured at the same sites

I want to test for differences in the abundance (counts) of two types of lizards. Both types were sampled at the same sites (n=284). The distribution of both types is overloaded with zeros and lizards ...
0
votes
0answers
45 views

What can go wrong if I use normal 2sls IV regression for count data?

My dependent variable is count data and for explanatory analysis we used a negative binomial regression. One variable is endogenous and we would like to instrument it. However, in Stata there is only ...
1
vote
1answer
37 views

Running Latent Dirichlet Allocation (LDA) on word counts

I have difficulties understanding the VB implementation lda-c. In particular, the method expects as input a bag-of-words representation of documents, where distinct words appearing in a document are ...
1
vote
0answers
48 views

How can I model such count data with underdispersion

Here is an example of my data: 2 6 4 5 2 5 4 4 2 3 3 5 5 6 5 6 15 19 16 9 14 14 11 10 6 4 2 In my assumption, the sequence can be separated into regimes, for instance, 2 6 4 5..6 6 5 6 belong to the ...
0
votes
0answers
26 views

What method is appropriate? - analysing patterns of attacks by region

I am currently working on a project using a dataset on global terrorist attacks. For each attack there are details on the type of attack which has 7 levels (e.g. bombing, assasination etc.), the ...
2
votes
1answer
50 views

How to categorize count data

I have count data (basically histone modification data). The counts represents number of reads falling into each genomic region. I need to divide my data into $\text{low|medium|high}$ based on the ...
0
votes
0answers
26 views

Quantifying 1 factor count data with variable # of categories

I feel like this should be an easy question, but after lots of looking I"m stumped. I deal with continuous data almost exclusively, but in my infinite wisdom I designed this study that yielded ...
0
votes
0answers
49 views

Can I use SAS Copula procedure or Matlab copulafit to fit count data (Poisson or Negative Binomial)

I want to simulate data (x,y) with dependency using copula. Matlab has the function for t copula, Clayton, Frank, and Gumbel bivariate Archimedean copulas. But I am not sure if these functions could ...
0
votes
0answers
20 views

Which reliability index (if any) is appropriate?

Subjects were asked to endorse whether or not they have experienced various negative events (yes/no). I want to use a measure that is a sum of how many events they have experienced. Since this measure ...
0
votes
0answers
77 views

Which model for under-dispersed data? Effect of poverty on adaptive capacity

I am working on the specific effects of poverty on farmers' adaptive capacity in sub-saharan Africa (N=1211). In bivariate, two of my poverty measures have an non-linear relationship with adaptive ...
0
votes
0answers
17 views

resources for temporal count data

I am currently in the middle of analyzing count data. The count data is gathered once per day for more than 100 days. There are two populations that are counted everyday. Is this type of count data ...
0
votes
0answers
78 views

Selecting best count model using Stata's countfit?

I am using the countfit command in Stata 13 to determine which count regression model is the best fit for the a model of number of current domestic migrants, with independent variables: workers in ...
0
votes
0answers
19 views

Using count data with number of days

I have two populations A and B. The data consists of count data per number of days after an event has occurred. For example: ...
0
votes
0answers
71 views

Multiple comparison for model fit with glmer

I would like to use glmer to fit a model of seed count data, and then compare the means from each plant line at two different temperature regimes. I am really not sure how to go about doing the ...
1
vote
0answers
50 views

Estimating confidence intervals on count data from classifier output on varying populations

I have a widget factory which produces a varying number of widgets each day. Every day, I need to report the number of faulty widgets. But the proportion of faulty widgets is expected to change every ...
2
votes
0answers
28 views

Identify deviations from trend

I have a panel with individuals $i$ measured at time $t$ (the date). The outcome is a count, which we can call $C$. There's some general time trend $g$ affecting all individuals. $E[C_{it}] = ...
2
votes
1answer
48 views

Exploring effect of treatment on count data

I've collected data on animal visitation at four different points in time. The four time points represent the total animal visitations over a three day period, i.e. 3 days of monitoring at four ...
1
vote
1answer
47 views

Count data and heteroscedasticity

Why are count data characterized by heteroscedasticity? If this a violation of the main linear models' assumptions of homoscedasticity, does it mean that in the relevent models for count data ...