Elkhart Group Ltd. can help you with almost any standard analysis you have in addition to more custom and specific analyses.
Our statistical consultants can help you conduct independent samples t-tests to compare means between two groups or paired samples t-tests for pre-post measures on the same people. If you have more than two groups, we can help running ANOVAs for any number of groups and factorial designs for experiments. Frequencies can be compared using chi-squared tests as well as Fisher exact tests for small sample sizes.
Many statistical tests assume a continuous or normal distribution, but not all data are like that. Two compare whether an outcome is higher in one group than another, we can help you use the Wilcoxon rank-sum test or for paired and repeated measures data the Wilcoxon signed-rank test. Multiple groups can be compared using the Kruskal-Wallis test. If you want to compare whether two samples come from the same distribution, we can use the Kolmogorov–Smirnov (KS) test.
We can help you assess the strength of the relationship between two variables using correlations or if you want to see if a relationship holds controlling for other variables, linear regression. Binary and ordered outcomes such as whether a patient has a disease or the stage of the disease can be analyzed using logistic or probit regression and count outcomes such as the number of accidents or falls someone has using Poisson regression.
Whether you just need to see your data and understand what they are saying, or need to visualize a statistical summary of a model; are using the graph internally or need publication and presentation quality graphs, our statistical consultants can help you show off your data and findings and make graphs that meet any standards. Most graphs can be exported to a wide variety of file formats so you can modify them, and we can output at any resolution so you have high quality images no matter how big you end up displaying them.
Statistical analyses come with assumptions. If those assumptions are not true, they can affect the validity of the test results. If you have data you worry may violate assumptions, or a decision that you cannot risk being wrong, just let the statistical consultant assigned to you know. Our experts at Elkhart Group Ltd. will run sensitivity analyses using multiple test statistics that make different assumptions. We can use parametric and non-parametric analyses, bootstrapping, and Monte Carlo simulation. We will summarize the results and let you know if your finding held up under all tests, and if not, under which circumstances the effect went away.
Premier: Our most bespoke offering. Let our consultants create solutions and reports that turn your data into strategic insight.
Academic: For accredited, not-for-profit universities. From methods write up assistance to advanced model analysis, we are experienced researchers who enjoy scientific discovery.
© 2013 - 2017 by Elkhart Group LLC