Variability Edge(X) Notes: Module 1 Error Variability: o Measurement process may be prone to error, giving measurement variability. Difficult to distinguish between measurement and natural variability, hence collectively these factors are error variability. o Necessary to replicate experiments, multiple observations give more accurate information Group Variability: o Where male and female forearms are measured we find differences in observations because groups tend to have different lengths. o Cannot conclusively claim, "males have longer forearm's than females" - not universally true o Instead talk in averages, claim "average male forearm length is longer than females". Sampling Variability: o The average lengths of forearm's calculated depends on sample. Quantitative Variables: o Quantitative variables represent measurements e.g. height or temperature. o Often continuous, taking any value over some range. o Discrete variables have a small number of possibilities, such as an age measured in years Categorical Variables: o Variables that can take on one of a limited, usually fixed, number of values, assigning each individual or other unit of observation to a particular group or nominal category o Nominal variables are categories with no order between them. Ordinal variables have order Designing Studies (Information more relevant to Research project) Comparative Experiments o Comparative experiment - two samples exposed to different conditions are compared. o The sample size will depend on the effect we're trying to detect, as well as other factors. Randomisation plays a fundamental role in avoiding bias. o Best way to avoid placebo effect within a study is to use a blind study design, subjects unaware of the treatment they are receiving. o In experiments where the assessment has a subjective element, a double blinded study is recommended - participants and experimenters unaware who is receiving which treatment. o A scientific control group allows researchers to minimize the effect of all variables except the independent variable. The control group, receiving no intervention, is used as a baseline to compare groups and assess the effect of that intervention. Observational Studies: o Has same aim as an experiment but is passive, often working with existing data such as medical records. Difficult to establish causation.
Hypothesis Testing: Randomisation test - statistical hypothesis testing to eliminate chance of chance variability. o Null hypothesis denoted as H0. States exact opposite of what an investigator or an experimenter predicts or expects. o Purpose is to prove whether test is supported, which is separated from the investigator's own values and decisions. o Assuming the null hypothesis is true, we calculate the probability that we could get data like what we saw just by chance. This probability is called the P-value of the test. P-value Small suggests that null hypothesis might be wrong, giving evidence for the second explanation. Large means data could have happened by chance, so we say it gives inconclusive evidence of an effect. If P-value is less than 0.Oft we say "the results were significant at the ft% level". This will also often be written in journal articles as "the results were found to be