Statistical Consulting Services

Sample Projects

1.  Categorical Data Analysis

The study investigated the potential benefits and risks of the use of non-vitamin and non-mineral (NVNM) supplements. The following research questions were analyzed: (1) whether there are associations between the use of NVNM supplements and the income level and the ethnic background; (2) whether NVNM supplement users with certain medical condition alert a health care professional. All data are categorical. To answer these questions, contingency tables were constructed and Fisher’s exact tests were used to test whether the associations are significant. The logistic regression models were built to adjust for other covariates to validate the significance of the associations.

2.  Analysis of Variance

The purpose of the experiment was to determine if core strength is correlated to running speed in high school athletes. In the study, one group of athletes was exposed to a core strengthening program and one control group was not. The pre- and post-test measurements were taken from each group. The data on confounding variables were also collected. An analysis of covariance model was used to determine the effect of the core strengthening program while controlling for the confounders.

3.  Regression Analysis

The research objective was to assess the influence of personality hardiness, stress, and psychological resource on career transition outcomes, in particular, the length of time it takes to complete the transition. The personality hardiness, stress, and psychological resource are measured by standard instruments from the existing literature. These instruments were validated in the analysis. The response variable was appropriately transformed such that it was approximately normally distributed. The linear regression model was built to assess the effects of the above three factors while adjusting for other relevant covariates.

4.  Linear Mixed Model

The goal of the study was to determine whether the consolidation of Illinois school districts resulted in a change in the discipline practices. The discipline practices were measured by the expenditure, transportation cost, and students’ reading and mathematics scores. They were measured repeatedly for several years from each school district before and after consolidation. To account for the correlation of measurements within each school district and other factors that might affect these measurements, the linear mixed model was adopted to determine whether there was a significant impact of consolidation on these discipline practices.

5.  Non-linear Mixed Model

The project analyzed the maternal care of two species of monkey: Callimico goeldii and Callithrix jacchus. Callimico goeldii gives birth to single offspring while Callithrix jacchus gives birth twins. It was conjectured that Callimico goeldii, with its single births, has shifted its maternal care to increase behavioral care and the amount time dedicated to infant care relative to Callithrix jacchus. This study measured a variety of maternal care to confirm this hypothesis. Each subject in the study was measured repeatedly for several weeks and the response variables are count data. The non-linear mixed models were applied on the data to establish whether the two species are different in terms of maternal care or not.

 

For a list of some of the other projects we have worked on, please click here.