About Stat59

Stat59 was developed by a physican researcher with a Masters degree in Statistics and extensive experience in web development and statistical computing. Working as a statistcal consultant, the developer found that several patterns in research development were common, and often led to waste of time and money during the research process.


The Stat59 Team

Jeffrey Franc, MD, MS, MSc, FCFP.EM, Dip Sport Med.

Founder and Chief Technical Officer

Dr. Franc is a Clinical Professor of Emergency Medicine and holds an MS in Applied Statistics from the Rochester Institute of Technology. He brings over 20 years of research experience and is a senior statistician at an international research center in disaster and emergency medicine. He is the author of 20 peer reviewed journal articles, 10 oral abstracts, and 20 poster presentations. In addition he has given over 200 invited presentations teaching in research methodology, statistics, disaster medicine, emergency medicine, and sport medicine.

Dr. Franc has nearly twenty years of experience as a software developer in R, php, java, C, and python, and has been a Linux system administrator and web site administrator since 1994. He is the principle architect of several software packages and apps.

Manuela Verde

Chief Operating Officer

Dr Verde is a specialist in Anesthesia and Intensive care and holds a Masters Degree in Disaster Medicine. She brings extensive experience with data entry, collection, and cleaning.


Why Stat59?

Stat59 was developed to address the five most commonly seen opportunities for improvement:

1. Errors in Initial Study Design

Inital errors in study design are common. Without understanding how the data will be analyzed it is very difficult to understand which data to collect and how to collect it. Even small errors, like failure to randomize the subjects properly, can weaken the validty of a study and prevent its publication. Furthermore, many errors in study design simply cannot be corrected later, leading to a tremendous waste of time and money. At Stat59 we will work with your team from the beginning to ensure that the initial study design is the best possible design for your research project

2. Data Tabulation is Haphazard

Well meaning researchers often tabulate their data on spreadsheets. Many go so far as to even total the columns, and calculate basic statistics such as mean and standard deviation. Unfortunately, such spreadsheets are entirely unsuitable for more sophisticated statistical analysis. Furthermore, errors in these spreadsheets are easy to commit and hard to find as the spreadsheets contain no way to ensure data is valid

The Stat59 web platform will provide your research team an easy to use web interface for data entry. Data entry is seamless, and validated in real time to minimize errors. The web platform can be accessed simultaneously by any number of researchers from your team, without the need to send a single spreadsheet back and forth between members.

3. The Wrong Test Statistic is Chosen

Purchasing traditional statistics software like SAS, minitab, or SPSS is similar to purchasing a bronchoscope. It is a useful tool, but can be very harmful if used without training. Statistics software is designed primarily for statisticians, and assumes that the user knows which test to use. As the software usually does not know the origin of the data or the design of your study, it will happily perform the wrong statistical test and give the wrong answers. Common mistakes are such things as calculating the mean of categorical variable, calculating the average postal code, or using Fisher's exact test when the basic test assumptions are not met.

Choosing the wrong statistical test is often not discovered until the peer review process. When the statistical test needs to be changed it may involve repeating the analysis. As the analysis will lead to different numbers and often different conclusions, it may lead to a very time consuming rewrite of the manuscript. The Stat59 platform is designed to prevent the wrong test from being done by matching data models carefully to statistical tests, and by careful asssessment of the study design

4. Translating the Statistics to a Narrative is Difficult

For many researchers who have not written a multitude of papers, it can be difficult to translate their ideas and numbers into a narrative manuscript. Furthermore, many written guidelines for manuscript prepartion exist, but are often specific to certain study designs.

ManuBot - a manuscript templating engine - will analyze your study design and data, correlate them with existing publication guidelines, and generate a manuscript template. Simply cut-and-paste the template into your usual manuscript editing software or word processor, and the template will guide through the process. In most cases the ManuBot will write a large portion of the methods and include references for the statistical methods used.

5. The Peer Review Process is Demanding

For certain peer reviewers of most journals are becoming more and more sophisticated. It is not common to see sophisticated comments such as:

  • 'Multiple p-values were obtained, how was the multiplicity accounted for?'
  • 'The null and alternative hypothesis were not properly stated.'
  • 'The basic assumption of the statistical tests used were not met.'
  • 'The authors did not confirm that the linear assumption was met prior to fitting the linear regression.'

ManuBot will analyze your manuscript and look for common causes of concern for reviewers. ManuBot will issue a warning on our manuscript template and tell you how to avoid that particular reviewer comment and how to respond to the comment if it should come back to you.

If any of these issues seem familiar to your team, the Stat59 solution can help move your team foreward.

How Can Stat59 Help?

Saves Time

  • Proper study design saves costly revisions
  • Data is cleaned and validated in real time. Prevents need for lengthy data reentry
  • Changes to data and analysis are available in real time. When you change the data the analysis is updated immediately. No need to wait for response from the analyst.
  • Time to write the manuscript can be greatly reduced by the Stat59 ManuBot templating engine.
  • The ManuBot anticipates reviewer comments and demonstrates ways to avoid them, speeding up the peer-review to revision cycle
  • The Stat59 interface if user friendly and needs no training, eliminating the need to learn complicated statistics software

Saves Money

  • Stat59 project plans cost a fraction of the cost of hiring a statistics consultant
  • No additional software costs. The Stat59 platform eliminates the need for statistics software packages, many of which are very costly
  • Prevents the need to hire an on-staff statistician
  • Allows research staff to spend their time on more productive tasks

The Research Cycle

Most research follows a similar cycle:
  1. Idea Generation
  2. Hypothesis Forming
  3. Creation of a Study Design
  4. Collecting Data
  5. Tabulating and Cleaning the Data
  6. Statistical Analysis
  7. Translating the Statistics to a Narrative
    • Summary Statistics
    • Hypothesis testing and effect measurement
    • Plots
    • Idea Generation
  8. Submission for Peer Review
  9. Responding to Reviewers Comments
  10. Publication
The Stat59 solution can help at all phases of the cycle. Our goal is to allow your research team to work on its strengths, and allow us to do the rest.

System Requirements

Stat59 is an all-in-one software solution that includes all phases of study design and data handling. Stat59 is a web based software and requires only a web browser and internet connection. Stat59 can be launched from a desktop computer, laptop computer, tablet, or smartphone.


Technical Details

Stat59 statistical calculations are performed using R: A Language and Environment for Statistical Computing. Data are stored on a MySQL server.