Establishing hypotheses Flashcards
What is statistical model?
a simplified mathematical representation of data generating process
What should a good hypothesis be?
Justifiable & interpretable
Quantifiable through data collection
Considered valuable
What are the 3 types of random variable?
Continuous (weight, height)
Discrete (age, counts)
Categorical (sex)
what is the response variable?
the thing we wish to predict
what is the explanatory variable?
the variable we wish to use for prediction
What 2 components should be thought about when looking at data
Systematic component - average/typical behaviour of response variable
Random component - characteristics of distribution due to other factors
what is a null hypothesis?
a hypothesis that attaches a unique value to parameter concerned
(states there no sig dif in results so there is no causality)
what is the alternative hypothesis
the hypothesis of interest (differentiate it from the null)
How are bias & confounding avoided/minimise in study design?
randomisation as the whole population cannot be assessed
Clinically and statistically significant difference
Clinically sig - effect sizes
Statistically sig - confidence intervals / hypothesis tests
what is a stochastic process and how does it effect test results
a process appears to vary in random manner eg heart attacks
there will be a distribution of outcomes (know as sampling distribution)
what does P value show
how likely you are to observe a test statistic that is more extreme than the observed value if the null hypothesis was true
If the P value is below statistical significance level what can we do?
reject the null hypothesis in favour of the alternative hypothesis
