[Zishi]

Question:


I'm facing problem with part (iv) of this question. To find the mean, I have to change the "No. of pupils in a school" to "class boundaries", find the midpoints and then use the midpoints to find the mean. Here is the table I have drawn which shows "class boundaries" and frequencies:

Isn't it right?

If yes, then shouldn't the midpoints be 50, 125, 175, 225, 300, 400, 525?

Mark scheme uses 50.5, 125.5, 175.5, 225.5, 300.5, 400.5, 525.5 as midpoints. . .

. . .where am I going wrong? Any help would be much appreciated.

[ghostwalker]

(Original post by Zishi)
Isn't it right?

If yes, then shouldn't the midpoints be 50, 125, 175, 225, 300, 400, 525?

Mark scheme uses 50.5, 125.5, 175.5, 225.5, 300.5, 400.5, 525.5 as midpoints. . .

. . .where am I going wrong? Any help would be much appreciated.

You're dealing with discrete data - you can't get half a pupil!

So your frequency classes of <= 200, <=250 are
101 <= x <= 200, 201 <= x <= 250, for example.

Hence the class boundary is 200.5, etc.

[Zishi]

(Original post by ghostwalker)
You're dealing with discrete data - you can't get half a pupil!

So your frequency classes of <= 200, <=250 are
101 <= x <= 200, 201 <= x <= 250, for example.

Hence the class boundary is 200.5, etc.

Alright, so class boundaries are always in decimals for discrete data, right?

[ghostwalker]

(Original post by Zishi)
Alright, so class boundaries are always in decimals for discrete data, right?

Not necessarily.

Consider a cumulative frequency table for the value of 10p coins which 100 people have in their pockets.

Classes might be 0 <= x <=20, 30, <=x<= 50, etc. with a class boundary of 25.

[Zishi]

(Original post by ghostwalker)
Not necessarily.

Consider a cumulative frequency table for the value of 10p coins which 100 people have in their pockets.

Classes might be 0 <= x <=20, 30, <=x<= 50, etc. with a class boundary of 25.

Alright, that perfectly makes sense. Thanks a lot! (PRSOM)

[ghostwalker]

(Original post by Zishi)
Alright, that perfectly makes sense. Thanks a lot! (PRSOM)

You're welcome.

[Zishi]

(Original post by ghostwalker)
You're welcome.

Just to confirm it - for part (iv) of the following question, I think I need to get class boundaries:


So should they be 19.5, 29.5, 39.5, 44.5, 49.5, 59.5 and 69.5?

[ghostwalker]

(Original post by Zishi)
Just to confirm it - for part (iv) of the following question, I think I need to get class boundaries:


So should they be 19.5, 29.5, 39.5, 44.5, 49.5, 59.5 and 69.5?

The data here would be continuous since we're talking about a weight, so I'd expact the boundaries to be 20, 30,...

[Zishi]

(Original post by ghostwalker)
The data here would be continuous since we're talking about a weight, so I'd expact the boundaries to be 20, 30,...

Ahh, my bad! Okay, so if we were talking about people, my class boundaries would have been correct then?

[ghostwalker]

(Original post by Zishi)
Ahh, my bad! Okay, so if we were talking about people, my class boundaries would have been correct then?

If, for example, it was the number of people in classroom, then yes, your class boundaries would be correct.

[Zishi]

(Original post by ghostwalker)
If, for example, it was the number of people in classroom, then yes, your class boundaries would be correct.

Hmm, thanks again.