Notes in l. EBM: Biostatistics

To Subscribe, use this Key

Status	Last Update	Fields
Published	07/28/2024	{{c1::Independent variables}} is the thing/ things that are changed
Published	07/28/2024	{{c1::dependent variables::what variable type}} is the result that is monitored
Published	07/28/2024	Independent variables can have {{c1::levels}}.
Published	07/28/2024	{{c1::categorical}} variables are discrete
Published	07/28/2024	{{c1::ordinal }} variables have rank (eg. 1st, 2nd. high risk low risk)
Published	07/28/2024	{{c1::nominal}} variables have no rank. {{c1::dichotomous}} variables are a sub category that fall under a binary
Published	07/28/2024	{{c1::continuous variables}} can take any value, typically expressed as an interval value or ratio
Published	07/28/2024	{{c1::Descriptive statistics}} {{c2::summarize and organizes data}}
Published	07/28/2024	{{c1::measures of central tendency}} include mean, median, mode, IQR, range. These are best used for {{c2::continuous variables}}
Published	07/28/2024	{{c1::distribution graphs}} are used for {{c2::continuous variables}} - they can reveal if a data set is {{c3::skewed, normal, bimodal etc.::poss…
Published	07/28/2024	{{c1::measures of variability graphs}} are used for {{c2::continuous variables}} - this includes standard deviation and standard error
Published	07/28/2024	Standard error is a {{c2::statistic}} that provides {{c1::estimate of standard deviation parameter}}
Published	07/28/2024	Standard error is a {{c2::statistic}} that tells us {{c1::how different the parameter is estimated to be}}
Published	07/28/2024	Categorical variables are normally shown {{c1::graphically - eg pie chart bar graph heatmap}}. Also tend to be described using {{c2::percent…
Published	07/28/2024	When making inferences we have {{c1::parametric and non-parametric tests}}
Published	07/28/2024	Parametric tests require {{c1:: normal distribution::distribution type}}, {{c1::quantitative data::qual/quant}}, {{c2::continous variables}}…
Published	07/28/2024	Para non para tests for comparing two samples same pop
Published	07/28/2024	Para non para tests for comparing measures from single sample
Published	07/28/2024	Para non para tests for comparing >2 samples
Published	07/28/2024	Para non para tests for linear association
Published	07/28/2024	Para non para tests for when there are more than one independent variable
Published	07/28/2024	Risk - {{c1::measure of association between exposure and outcome}}Equation: {{c2::a/a+b ; c/c+d}}
Published	07/28/2024	Absolute Risk reduction (risk difference) = {{c1::c/c+d - a/a+b}} is the difference in the risk of the outcome in the control group and the …
Published	07/28/2024	if Absolute risk reduction (risk difference) = 0, {{c1::risk is same for experimental and control}}
Published	07/28/2024	if absolute risk reduction < 0, {{c1::risk is increased in exposed group}}
Published	07/28/2024	Number needed to treat = {{c1::1/ absolute risk reduction (risk difference)}} = {{c2::number needed to achieve an additional good outcome}}
Published	07/28/2024	Relative Risk (ratio) = {{c1::a(c+d)/c(a+b)}}. A ratio of risks. Often referred to as the risk ratio. This is the ratio of the risk of the…
Published	07/28/2024	Relative risk reduction is {{c1::the estimate of baseline risk removed (or added) by an intervention.}} = {{c2::1 - relative risk}}
Published	07/28/2024	Odds – {{c1::essentially probability of having the + outcome, ratio of outcome of interest to not outcome}} = {{c2::a/b ; c/d}}
Published	07/28/2024	Odds ratio – {{c1::odds of outcome with exposure / odds of outcome without exposure}}
Published	07/28/2024	Odds are often seen in {{c1::case control studies:: study type}}. Relative risk is often seen in {{c2::cohort study or RTC}}
Published	07/28/2024	p-value is {{c1::the probability of getting your statistic from the population assuming the null hypothesis is true}}
Published	07/28/2024	type 1 error is {{c1::a false positive}}. It is equivalent to {{c2::the significance, alpha}}
Published	07/28/2024	type 2 error is {{c1::a false negative}}. It is directly related to {{c2::the power of a study, 1- beta}}
Published	07/28/2024	Study power depends on {{c1::sample size, alpha, amount of scatter, etc.}} Power determines the {{c2::minimal detectable difference}} that w…
Published	07/28/2024	CI {{c1::provides a plausible range of likely differences that captures the population parameter}}
Status	Last Update	Fields