STATS 10Xs

Question	Answer
What is an artefact?	An untrue pattern of data caused by fault data collection processes
What is an Association?	An association is a relationship, in which two variables are dependent on each other. Such as length of an iron beam and it's heat. It's unlikely that a pattern will show with out association.
What is also the centre of a graph?	This is referred to the median of the data set, and represents where 505 of the data is on one side of that point, and visa versa.
What is a Cluster?	Where there is a lot of data surrounding on point.
What is an Entity?	An entity is a set of data, or a type of variable such as height. What is a
What is an Estimate?	A number calculated from the data, used to guess an unknown parameter.
What is modality?	Modality refers to the reoccurring nature of high frequent sets of data. i.e Bimodal.
What is a nominal variable?	A categorical variable where there is no order
What is an Oddity?	An oddity is where there is an unusual set of data.
What is an Ordinal variable?	A categorical variable that has order, such as Red, Orange, Yellow, Green, Blue, etc
What is the difference between overprinting and overlapping?	overprinting: A problem with scatter plots in which a data point is printed on top of another. Overlapping: This is a visual notion in which points go close to each other.
What is Risk?	The way of expressing the chance of something occurring.
What is absolute risk?	The chance of a person in a population will have a specified event (Usually percentage)
What is relative Risk?	The comparison of the risk of a particular event with the risk of another event.
Study Standard Deviation	LOL just do it.
What is a Systematic Bias?	Consistent bias caused by the system or process of collecting data.
What are the points of an observational study?	The person doesn't change anything. We study the effect of a factor of interest in units (Usually individuals) Compare the units with what happened Comparisons are meant to be fair. Random sampling.
What is a cross sectional study and a longitudinal study?	Cross-section is where it's an observation of a snapshot in time. You know Longitudinal.
What can an observational study say about something?	They can identify probably causes for effects but CANNOTreliably establish causation.
What can an experiment say about something	Assuming that the experiment was well designed and executed, you CAN say there there is reliable established causation.
What are the points of an experiment?	Observe the effects of a treatment measure the response of the treatment Compare the different treatment ranges, especially with the control group.
What are the different types of experiment?	Completely randomised design. Randomised block design
Go to page 6 and study it chapter 2	lmao do it.
What does it mean by "Under chance acting alone"?	It means that it is possible for the data to be made because of random entries and that there is no association or relationship of the variables.
What is it called when people are put into a group based by some factor, and then randomly given treatments?	Randomised Block Design
What are the two types of errors that can occur with sampling?	Sampling, and non-sampling
An issue where any person can take part is a form of what survey problem?	Non-sampling, self-selection bias. This can cause issues as it may not give an accurate representation of the population
What is the name of the issue where you do not select an appropriate sample of a population?	Selection Bias, as it will not give good data based on the population.
What is the difference between question effects and behavioural considerations?	Behaviour: People answer with an answer that is more socially desirable Question effects is where it loads a socially desirable answer with intent or without.
What is the problem where you take a survey of a sample of a population, and that sample doesn't acuratlly depict the population you wish to talk about?	This is a transfer finding non-sampling error.
What is the issue where the question order, layout, interviewed by phone, person or by email cause with a survey?	This is called a survey format effect. An example of this is where surveys are given out via email but offer the prize of an iPad. And people don't care about the results, just the iPad.
What type of data would be used with: Dot plot Stem and leaf Box plot Histogram	Numerical
Why would use use a Histogram over a Barchart?	A histrogram is the same data set and is used like a dotplot, to show along the x axis how many times a value was with in that range. A box plot can compare multiple variables.
What is a way to compare two numeric values?	A scatter plot Tile Density plot
What is a bar chart used for?	Categorical values!
What are the two factors in calculating Standard Error	Standard deviation of the sample and sample size
Do get a n more times precision with a study you need what of a sample size to get that.	n^2
There are three situations a, b, c for difference of proportions. What are they?	A: Proportions of two independent samples B: One sample size of n, multiple responses. C: One sample size of n, many y/n responses.
What is it called when random people are put into random treatments	Randomised experiment!
What is it called when there is a numeric variable, but it doesn't stop?	Continuous
What is it called when there is a numeric variable and it counts , 1, 2, 3,	Discrete
What is it called when there is a Categorical variable and it doesn't have an order	NOMINAL! :D
What is it called when there is a categorical variable, that has an order?	ordinal

Next up

Description

Resource summary

Similar

	Created by Dan Richards over 7 years ago