-
The irony of exclusionary inclusion: Food for thought on the inclusive collection of sexual orientation and gender identity information and implications for data analysis
What is your gender? Man? Woman? Do you consider yourself to be heterosexual or homosexual? Society has come a long way from this kind of binary categorisation. Thankfully, research has also come a long way in asking questions about sexual orientation and gender identity that are more inclusive and attempt to recognise both the continuum […]
-
Enslaved to Model Fit? We have nothing to lose but our chains.
One of the major group of indices that a statistician works with are those examining model fit. Model fit statistics provide us an indication of the degree to which a model we have specified fits the data at hand. Such statistics are available for a whole gamut of models from the most basic to the […]
-
Outliers are people too you know
In traditional undergraduate statistics courses we are generally taught that outliers are bad news. Outliers are pesky peeps who don’t conform. They stand out from the pack. We should banish them from our data set and be done with them. However there are a number of points to consider about our outlying data points that […]
-
To Bonferroni or not to Bonferroni.. that is the question
Hi Dr S, I just have a quick question. I have run some tests and when I apply Bonferroni adjustments to them I lose all my significant results. Do I have to apply the Bonferroni? Can I get away with not using it? Kind regards Bob A. Student Before we dive head first into this […]
-
We are non-significant results, hear us roar
I conducted a series of meta-analyses for my PhD. I earnestly read up on meta-analytic techniques in a range of text books. In the process I came across the notion of publication bias. Publication bias suggests that a manuscript is more likely to be accepted for publication by a journal if it reports on statistically […]
statisticat
Musings from the world of research methodology and statistics
-
The irony of exclusionary inclusion: Food for thought on the inclusive collection of sexual orientation and gender identity information and implications for data analysis
What is your gender? Man? Woman? Do you consider yourself to be heterosexual or homosexual? Society has come a long way from this kind of binary categorisation. Thankfully, research has also come a long way in asking questions about sexual orientation and gender identity that are more inclusive and attempt to recognise both the continuum…
-
Enslaved to Model Fit? We have nothing to lose but our chains.
One of the major group of indices that a statistician works with are those examining model fit. Model fit statistics provide us an indication of the degree to which a model we have specified fits the data at hand. Such statistics are available for a whole gamut of models from the most basic to the…
-
Outliers are people too you know
In traditional undergraduate statistics courses we are generally taught that outliers are bad news. Outliers are pesky peeps who don’t conform. They stand out from the pack. We should banish them from our data set and be done with them. However there are a number of points to consider about our outlying data points that…
-
To Bonferroni or not to Bonferroni.. that is the question
Hi Dr S, I just have a quick question. I have run some tests and when I apply Bonferroni adjustments to them I lose all my significant results. Do I have to apply the Bonferroni? Can I get away with not using it? Kind regards Bob A. Student Before we dive head first into this…
-
We are non-significant results, hear us roar
I conducted a series of meta-analyses for my PhD. I earnestly read up on meta-analytic techniques in a range of text books. In the process I came across the notion of publication bias. Publication bias suggests that a manuscript is more likely to be accepted for publication by a journal if it reports on statistically…
statisticat 