Simpson's Rule OCR MEI Numerical Methods

I've just been entirely defeated by an NM past paper question (Jan 2007 Q7).

I've found the wrong Simpson's estimates. Rather than finding them for h = 0.5 and h = 0.25, I have seemingly found them for h = 0.25 and h = 0.125 (respectively) instead.

Having checked my textbook and the specification to work out why, they contradict each other. The textbook says that Simpson's rule with n strips, $S_n$ is given by

$S_n = \frac{2M_n + T_n}{3}$

Where $T_n$ and $M_n$ are the trapezium and midpoint estimates.

The specification however says that

$S_{2n} = \frac{2M_n + T_n}{3}$

Where $M_n$ and $T_n$ are defined as before and $S_{2n}$ is the Simpson's estimate with 2n strips.

So what is the strip number in Simpson's Rule? Is it the number of strips used in the Midpoint and Trapezium rule to get that Simpson's estimate, or is it twice that number? The latter would give me the right answer but the former is what I was taught.

...

the former is correct. the mark scheme must be wrong

the former is correct. the mark scheme must be wrong

I'm confused as well now: this, this and this say it is $S_{2n}$ and the explanation given in the latter link makes complete sense.