Stats (1 Viewer)

[ZaoKai]

For stats does anyone know when to use continuity correction? Or do we even need to use it for the Hsc? Thanks.

 

[synthesisFR]

For stats does anyone know when to use continuity correction? Or do we even need to use it for the Hsc? Thanks.

i was told u never need to
like if u want to u can
but u dont need to and wont lose marks if u dont

 

[indeed]

Is continuity correction necessary to know (cause we never learnt it)? If so, could someone please explain it thanks

 

[user18181818]

no it's not in the syllabus it's just extra knowledge

 

[scaryshark09]

i was told u never need to
like if u want to u can
but u dont need to and wont lose marks if u dont

bruh i learnt it for nothing then??

 

[Luukas.2]

Continuity correction is needed (formally) when approximating a discrete distribution (like a binomial) with a continuous distribution.

Mathematically, it should always be included when making such an approximation, and leads to a more accurate result.

However, HSCially, solutions will be accepted with or without the correction. Presumably this means that the situations that can't be solved without the correction will not be examined,

 

[tywebb]

here's an example whose solution shows why continuity correction gets better answer

here is my solution to part ii and you can see compared to answers from methods 1 and 2, that continuity correction gets better approximation

 

[bruhmoment.]

This is what the Cambridge textbook says (Key Ideas) (page 800):

- A binomial distribution can be approximated by the normal distribution with the same mean and standard deviation.
- Thus the binomial distribution B (n, p) or Bin (n, p) can be approximated by the normal distribution N (np, npq) , where q = 1 − p.
- N (μ, σ^2) means the normal distribution with mean μ and variance σ2.
- When approximating, we treat the discrete binomial variable X as if it were a continuous normal variable.
- For small values of n, apply the continuity correction. This means integrating between half- intervals, corresponding to the boundaries of the cumulative frequency histogram. For example, to approximate P (X = 8, 9, 10, 11 or 12) , we treat X as a continuous normal variable and find P(7.5 ≤ X ≤ 12.5).

I'm a bit confused as to what a "small value of n" can be defined as. Like what would be the cutoff point?