For Researchers

Biostatistics Drop-in Sessions

Biostatistics Drop-in Sessions have finished for 2022, but will restart in February 2023.

Biostatistics Drop-in Sessions provide researchers with research and statistical support for their projects, from conception to dissemination. Support is available at any stage of the research process, for both established and novice researchers.

In the meantime, please contact Dr StellaMay Gwini, Biostatistician, with your queries.

 

Authorship

As the Biostatisticians work collaboratively with researchers, it is anticipated that more often than not the contribution made by a biostatistician may warrant co-authorship. All researchers are required to discuss the role of the biostatistician and clarify issues of authorship at the beginning of a collaboration. Barwon Health Guidelines for Determining Co-authorship for a Biostatistician and the Guidelines for Collaborative Research and Authorship may be consulted to facilitate this discussion.

Statistical software supported

• SPSS – for advice and guidance only
• Stata – this is the main and recommended statistical software. The software is freely available for use by all Barwon Health staff through a network license and can be connected to through Citrix. Hence, if you require access first ensure you have been granted access to Citrix and you can contact the Biostatistics office for permission to use Stata.
• R is also available through Citrix, although no support is available through the Biostatistics Service.

REDCap (for data collection and storage)

It is recommended that wherever possible, staff should use REDCap for data collection. For details on how to access REDCap, please contact Richard Larsen via email [email protected] or ph (03) 4215 3371.

Short Courses

Check again soon. If your team or department would benefit from a single or series of statistics presentations/workshops, contact Dr StellaMay Gwini.

Biostats Tips: A touch on Linear Regression

When examining data, one of the most common interests is to know whether two or more attributes are related, and regression is well suited for that. A regression is a mathematical representation of the relationship between a dependent variable and two or more independent variable. There are many types of regression models but the most commonly discussed are linear, logistic, Cox and Poisson regression; and the choice of model depends on the dependent variable.

When the dependent variable is continuous (e.g. age, weight, height), linear regression can be used to model its relationship with both continuous and categorical variables. Whilst the t-test and ANOVA can also be used for between-group comparisons of means, they have stringent assumptions that are often violated with real-life data. Linear regression on the other hand is less stringent and more flexible.

What are the steps for conducting linear regression?

• Identify the dependent variable and the independent variable(s).
• Identify the assumptions of linear regression. One assumption of linear regression is that the dependent variable has a linear relationship with the independent variable(s), but this may not be true. If an independent variable is continuous, consider drawing scatterplots to examine this linear relation and explore whether data transformations are warranted in order to achieve the linear relationship.
• Fit the regression model with preferred Statistical analysis software e.g. SPSS, Stata, R.
• Test whether the rest of the assumptions are satisfied.

Here some useful resources for an introduction:

• https://people.duke.edu/~rnau/regintro.htm
• https://www.sheffield.ac.uk/polopoly_fs/1.531434!/file/MASH_simple_linear_regression_SPSS.pdf
• https://www.youtube.com/watch?v=ZkjP5RJLQF4&list=PLIeGtxpvyG-LoKUpV0fSY8BGKIMIdmfCi

 

Biostats Tips: Six Terms that Mean Something Different Statistically and Colloquially

By Kim Love and Karen Grace-Martin 

Statistics terminology is confusing. Sometimes different terms are used to mean the same thing, often in different fields of application. Sometimes the same term is used to mean different things. And sometimes very similar terms are used to describe related but distinct statistical concepts. However, the terms that cause the most trouble are those with a different English colloquial and statistical meaning. This is particularly difficult because the definitions are often similar, if not exact.

Here are six of the most common terms:

1. Significance

You’re probably familiar with the difference between statistical significance, generally indicating a p-value that is below a threshold, and the colloquial meaning of large or important.

2. Odds

In everyday English, people use the terms Odds and Probability interchangeably. In statistics, they’re measuring the same general construct – how likely an event is to occur – on different scales. This difference in scales has a huge impact on how you interpret the value. Odds measure the probability (p) of an outcome relative to the probability that outcome doesn’t occur; i.e. p/(1-p).

3. Bias

In colloquial English, bias means prejudice. It’s bad. Bias isn’t always a good thing in statistics, but it doesn’t have that inherent value judgment.

In statistics it is a measure of the difference between the value of a population parameter and the theoretical mean value of a statistic that estimates that parameter. More often it comes from having an unrepresentative sample.

4. Correlation

In statistics, a correlation is a specific measurement. It is a measure of the direction and strength of association between two variables.

On the other hand, the colloquial definition is much broader, to indicate any connection, match, or co-occurrence between individual events.

5. Error

Colloquially, an error is a mistake.

Statistically speaking, an error is the difference between the measured value for one individual and the value predicted by a regression model. There’s no mistake involved here. Just variation.

6. Random

Statistically, a phenomenon is random if individual outcomes are uncertain, but there is nonetheless a regular distribution of outcomes in a large number of repetitions.

While this is one usage of random in everyday English, it also often means strange or unexpected.

Want more Biostats tips direct to your inbox? Click here to subscribe to StatsWise, the monthly newsletter produced by The Analysis Factor.

Biostats Tips: The Difference between Interaction and Association

By Karen Grace-Martin from The Analysis Factor

Often in research we seek to examine the relationship between two or more factors and it is really easy to mix up the concepts of association (as measured by correlation) and interaction.

Whether two variables are associated says nothing about whether they interact in their effect on a third variable. Likewise, if two variables interact, they may or may not be associated.

In statistics, these terms have different implications for the relationships among variables, particularly when the variables are predictors in a regression or analysis of variance (ANOVA) model.

The association between two variables means the values of one variable relate in some way to the values of the other. On the other hand, an interaction between two variables means the effect of one of those variables on a third variable is not constant—the effect differs at different values of the other.

For a more detailed description, with examples, read the full article by Karen here.