Plotting regression coefficients and other estimates in Stata

Graphical presentation of regression results has become increasingly popular in the scientific literature, as graphs are much easier to read than tables in many cases. In Stata such plots can be produced by the -marginsplot- command. However, while -marginsplot- is very versatile and flexible, it has two major limitations: it can only process results left behind by -margins- and it can only handle one set of results at the time. In this article I introduce a new command called -coefplot- that overcomes these limitations. It plots results from any estimation command and combines results from several models into a single graph. The default behavior of -coefplot- is to plot markers for coefficients and horizontal spikes for confidence intervals. However, -coefplot- can also produce various other types of graphs. The capabilities of -coefplot- are illustrated in this article using a series of examples.

COEFPLOT: Stata module to plot regression coefficients and other results

coefplot plots results from estimation commands or Stata matrices. Results from multiple models or matrices can be combined in a single graph. The default behavior of coefplot is to draw markers for coefficients and horizontal spikes for confidence intervals. However, coefplot can also produce various other types of graphs.

Plotting regression coefficients and other estimates

Graphical display of regression results has become increasingly popular in presentations and in scientific literature because graphs are often much easier to read than tables. Such plots can be produced in Stata by the marginsplot command (see [R] marginsplot). However, while marginsplot is versatile and flexible, it has two major limitations: it can only process results left behind by margins (see [R] margins), and it can handle only one set of results at a time. In this article, I introduce a new command called coefplot that overcomes these limitations. It plots results from any estimation command and combines results from several models into one graph. The default behavior of coefplot is to plot markers for coefficients and horizontal spikes for confidence intervals. However, coefplot can also produce other types of graphs. I illustrate the capabilities of coefplot by using a series of examples.

Diffusion-weighted MR imaging of upper abdominal organs: field strength and intervendor variability of apparent diffusion coefficients

PURPOSE To determine the variability of apparent diffusion coefficient (ADC) values in various anatomic regions in the upper abdomen measured with magnetic resonance (MR) systems from different vendors and with different field strengths. MATERIALS AND METHODS Ten healthy men (mean age, 36.6 years ± 7.7 [standard deviation]) gave written informed consent to participate in this prospective ethics committee-approved study. Diffusion-weighted (DW) MR imaging was performed in each subject with 1.5- and 3.0-T MR systems from each of three vendors at two institutions. Two readers independently measured ADC values in seven upper abdominal regions (left and right liver lobe, gallbladder, pancreas, spleen, and renal cortex and medulla). ADC values were tested for interobserver differences, as well as for differences related to field strength and vendor, with repeated-measures analysis of variance; coefficients of variation (CVs) and variance components were calculated. RESULTS Interreader agreement was excellent (intraclass coefficient, 0.876). ADC values were (77.5-88.8) ×10(-5) mm(2)/sec in the spleen and (250.6-278.5) ×10(-5) mm(2)/sec in the gallbladder. There were no significant differences between ADC values measured at 1.5 T and those measured at 3.0 T in any anatomic region (P >.10 for all). In two of seven regions at 1.5 T (left and right liver lobes, P < .023) and in four of seven regions at 3.0 T (left liver lobe, pancreas, and renal cortex and medulla, P < .008), intervendor differences were significant. CVs ranged from 7.0% to 27.1% depending on the anatomic location. CONCLUSION Despite significant intervendor differences in ADC values of various anatomic regions of the upper abdomen, ADC values of the gallbladder, pancreas, spleen, and kidney may be comparable between MR systems from different vendors and between different field strengths.

Significant Differences Characterise the Correlation Coefficients between Biocide and Antibiotic Susceptibility Profiles in Staphylococcus aureus.

There is a growing concern by regulatory authorities for the selection of antibiotic resistance caused by the use of biocidal products. We aimed to complete the detailed information on large surveys by investigating the relationship between biocide and antibiotic susceptibility profiles of a large number of Staphylococcus aureus isolates using four biocides and antibiotics commonly used in clinical practice. The minimal inhibitory concentration (MIC) for most clinically-relevant antibiotics was determined according to the standardized methodology for over 1600 clinical S. aureus isolates and compared to susceptibility profiles of benzalkonium chloride, chlorhexidine, triclosan, and sodium hypochlorite. The relationship between antibiotic and biocide susceptibility profiles was evaluated using non-linear correlations. The main outcome evidenced was an absence of any strong or moderate statistically significant correlation when susceptibilities of either triclosan or sodium hypochlorite were compared for any of the tested antibiotics. On the other hand, correlation coefficients for MICs of benzalkonium chloride and chlorhexidine were calculated above 0.4 for susceptibility to quinolones, beta-lactams, and also macrolides. Our data do not support any selective pressure for association between biocides and antibiotics resistance and furthermore do not allow for a defined risk evaluation for some of the compounds. Importantly, our data clearly indicate that there does not involve any risk of selection for antibiotic resistance for the compounds triclosan and sodium hypochlorite. These data hence infer that biocide selection for antibiotic resistance has had so far a less significant impact than feared.

A Bayesian approach to estimating the regression coefficients of a multinomial logit model

A Bayesian approach to estimation of the regression coefficients of a multinominal logit model with ordinal scale response categories is presented. A Monte Carlo method is used to construct the posterior distribution of the link function. The link function is treated as an arbitrary scalar function. Then the Gauss-Markov theorem is used to determine a function of the link which produces a random vector of coefficients. The posterior distribution of the random vector of coefficients is used to estimate the regression coefficients. The method described is referred to as a Bayesian generalized least square (BGLS) analysis. Two cases involving multinominal logit models are described. Case I involves a cumulative logit model and Case II involves a proportional-odds model. All inferences about the coefficients for both cases are described in terms of the posterior distribution of the regression coefficients. The results from the BGLS method are compared to maximum likelihood estimates of the regression coefficients. The BGLS method avoids the nonlinear problems encountered when estimating the regression coefficients of a generalized linear model. The method is not complex or computationally intensive. The BGLS method offers several advantages over Bayesian approaches. ^