In a previous post I muttered about how unhelpful sensitivity and specificity are to practicing clinicians, and how what we really want to know are the predictive values of a test. Remembering the Table Really diseased Really not diseased Test +ve A B 1.. A/(A+B) Test -ve C D 2.. C/(C+D) 5.. […]
Category: stats
Sensitivity and specificity
Sensitivity and specificity are those sorts of things that can really get knickers twisted up something rotten. They sound like something you should be able to understand, they get used as if you understand them, and then you realise … it’s not quite as you thought … Really diseased Really not diseased Test […]
StatsMiniBlog: Odds and Probabilities
There’s something that is frequently wittered about but the odds are you’ve never really been bothered enough to care if there’s a difference between ‘probability’ and ‘odds’ (like relative risk and odds ratios). There are great reasons for this. Coffee, beer, ‘Take Me Out’ or a crash call to labour ward are four, for example. […]
StatsMiniBlog: Subgroup or sensitivity analyses?
Perhaps becoming a little obscure, but there are some folk in the world who become concerned about undertaking analysis in systematic reviews. Some of these are described as “subgroup” analysis, others “sensitivity”. What’s the difference, and why? […]
StatsMiniBlog: Exact vs. approximate
You may well come across descriptions in the stats parts of papers that describe how the authors have derived their confidence intervals using an exact method. Sounds very good, doesn’t it? Precise to the most precicestness. And yet … sometimes an approximate confidence interval is better. You see, it all means what ‘exact’ exactly refers […]
StatsMiniBlog. Regression
Now, regression is a bad thing if we’re talking development. It might be any number of really difficult to pronounce neurological conditions, or severe psychological trauma, or abuse/neglect. It’s not going to be good. In statistics, it’s not quite the same. Regression is quite often a good thing. But what is it? […]
StatsMiniBlog: Bland Altman Plots
Measuring things is what we do lots of, and we often want to measure things with a new machine. New, faster, shinier, cheaper, less invasive or more colourful … but we are almost always sold it as being highly correlated with the reference standard (p<0.001). Think – what is this correlation and p-value telling us? […]
StatsMiniBlog: Type I and II errors
After reading the title, most people now feel vaguely nauseous. If you throw in alpha and beta, or worse α and β, then there’s a distinctly bilious taste. Don’t get sick, though. Take a deep breath and fall back on what you already know: […]
StatsMiniBlog: Propensity scores
Propensity scores are used mainly in observational studies assessing treatments as a way of balancing out measured variations in who received a treatment and who didn’t. In most observational studies, there are things which will have pushed the doc into prescribing the medicine in question, or the surgeon to take the knife to that patient […]
StatsMiniBlog: Bootstrapping
As mentioned in previous posts, part of the joys of playing with numbers is in making inferences about how the future will be. This is often why you’re looking for a confidence interval, showing (sort of) where the truth lies with 95% certainty. How you build these CI is interesting. One way is to assume […]