Registered  members with 100+ posts do not see Ads

Interview offers & dates 2016-17

Status
Not open for further replies.

biom

Regular Member
Supreme Overlord of the Chatbox
We don't know how Unsw marks interviews, whether by percentages out of 10 or 100 or whatever. But if true to the words they wrote to us the scores will be standardised (same as ATAR & UMAT) so doesn't really matter.

The estimated percentage given in the email is chance of getting a place not interview score. However see my explanation below I think we can *roughly* relate it to the percentile of one's interview score.



IMO you are absolutely spot on. Since Unsw already knows the place cutoff and your ATAR & UMAT, if coupled with max intv score they already can tell whether you are or are not getting a place. It doesn't make sense to say getting the maximum interview score gives you >X% chance.

I kinda think Unsw wrote like that to avoid any legal challenges. The real meaning of that is more like: If Unsw were to put 100 typical intv scores into a bag and draw one out for you, you have >X% chance it combines with your ATAR/UMAT to above the cutoff. Make sense? Thus I approximate it to the percentile of your interview score (i.e. >70% chance means 30%ile+ score).

(Writing this in a hurry, will edit if I can think of better ways to describe).
A1,
have a look at my reply earlier.

I don't think that you can link the likelihood to the percentile except by coincidence.

From the info posted what they are saying is.... given
1. your scores for UMAT and ATAR and
2. the distribution of scores that we have seen on interviews

it is x% likely that you interview score will be high enough to get an offer.

I think that this is a classical confidence interval exercise... happy to be corrected by someone that understand stats properly:confused:
 

A1

Rookie Doc
Moderator
A1, have a look at my reply earlier.
I don't think that you can link the likelihood to the percentile except by coincidence.

I'm not good enough with maths/stats to think more deeply into this^ haha.

You could well be right with your assertion. On my side I did emphasize the word *roughly*, to use the approximation to assist my explanation. I don't even understand "classical confidence interval" let alone using it to explain something else :p
 

pills

Member
I don't think that you can link the likelihood to the percentile except by coincidence.

From the info posted what they are saying is.... given
1. your scores for UMAT and ATAR and
2. the distribution of scores that we have seen on interviews

Given that they know nothing about your interview skills, after looking at the distribution and seeing the required interview z-score, the only reasonable way of converting it into a likelihood/percentage chance value would be, as you say, to use the sample of students that have already been interviewed and see what percentile rank your minimum interview z-score corresponds to.

In essence, I think you guys are using different forms of the same measure ;) Thanks for your replies from what I gather I should simply ignore the "assuming you will have achieved the maximum interview score" part of it cuz it doesn't really make sense to include it.

Also just in case anyone is curious I got a similar email with ">50%", so maybe that's as specific as it gets?
 

Mana

there are no stupid questions, only people
Administrator
Nope, nothing of the sort. I'm waiting for a1 and mana to explain

Mana and A1 we need you. Where are you when we need it the most? :)

Wasn't my turn on the MSO shift, sorry.


To clarify a few things from above: interviews, especially panel interviews, are scored out of a certain number (usually a nice round number like 100 is used but of course varies between universities). (MMI interviews have the possibility of particular stations being scaled up or down based on difficulty).

However, the important thing is what your *rank* in the interviews is compared to other people, rather than the raw number. That is, the mean mark for interview should be scaled to the 50%ile even if the score is actually 80/100. The same ranking criteria is applied to interview score as it is to UMAT and ATAR - they will plot everyone's UMAT and convert it with some statistical mumbo jumbo to a z-score, and then plot everyone's ATAR and convert that to a z-score, and plot everyone's interview score and convert that to a z-score. They will then aggregate those scores to a useful rank. To offer interviews, they can just use the aggregate of z-score from UMAT and ATAR. In that sense, biom's explanation above is a rather simplistic, but conceptually accurate, explanation.
 

shadowduke

Monash MD I
Received the email from UNSW this morning. It gave me a >50% chance, so I suppose that's decent (?), though it's provided the maximum interview score is achieved. I'm curious as to if anyone else received a higher estimate? Or whether they just tell you whether you have >50%, or <50% chance.
I was predicted >50% (95%/99.85) and so far the highest I've heard is >50% (97%/99.95). I believe this will be the highest that UNSW will predict, as they don't want to make you feel too comfortable coming into the interviews.
 

macander

Member
Is the 'chance' always quantified as '>50% or < (= :)) 50%' ?
Are they ever more specific?
One would envisage that there may be some (interstate) high-flyers :) with (ATAR/UMAT) combo of, say 99.95/99 that - with a perfect interview - will definitely exceed the (already known) 2st round cutoff -> their 'chance' would be, well, exactly 100% (which is still '>50%)
Has anybody (A1 ?) ever received their estimation of a 60, 70% etc 'chance' ?
 

biom

Regular Member
Supreme Overlord of the Chatbox
Wasn't my turn on the MSO shift, sorry.


To clarify a few things from above: interviews, especially panel interviews, are scored out of a certain number (usually a nice round number like 100 is used but of course varies between universities). (MMI interviews have the possibility of particular stations being scaled up or down based on difficulty).

However, the important thing is what your *rank* in the interviews is compared to other people, rather than the raw number. That is, the mean mark for interview should be scaled to the 50%ile even if the score is actually 80/100. The same ranking criteria is applied to interview score as it is to UMAT and ATAR - they will plot everyone's UMAT and convert it with some statistical mumbo jumbo to a z-score, and then plot everyone's ATAR and convert that to a z-score, and plot everyone's interview score and convert that to a z-score. They will then aggregate those scores to a useful rank. To offer interviews, they can just use the aggregate of z-score from UMAT and ATAR. In that sense, biom's explanation above is a rather simplistic, but conceptually accurate, explanation.
Mana,
I think that it is about the score not the ranking.

Imagine in my example that Person A was ranked 2nd in the interview and Person B was ranked first. Person A most likely - it depends on the actual score - gets the offer due to their superior scores in UMAT and ATAR.

ps - thanks for the "simplistic" comment. If anyone can look at these descriptions and say that then they are good descriptions!:)
 

A1

Rookie Doc
Moderator
I was predicted >50% (95%/99.85) and so far the highest I've heard is >50% (97%/99.95). I believe this will be the highest that UNSW will predict, as they don't want to make you feel too comfortable coming into the interviews.

On my own UNAPPROVED formula I gave myself (99.80/high-100%ile) better than 90% chance but yep, Unsw gave me the same >50%.

Heck with a max interview score I would be somewhere around 150% chance, so I already dismissed that "with max interview score" since last year :D
 

phosphene

Monash MD I
On my own UNAPPROVED formula I gave myself (99.80/high-100%ile) better than 90% chance but yep, Unsw gave me the same >50%.

Heck with a max interview score I would be somewhere around 150% chance, so I already dismissed that "with max interview score" since last year :D

Hahaha damn based A1 be slayin'.

Now I'm just hoping WSU does the same with its confirmation emails, though if I end up going to Syd for UNSW, may as well stay for WSU ones. d:
 

Registered  members with 100+ posts do not see Ads

Mana

there are no stupid questions, only people
Administrator
Mana,
I think that it is about the score not the ranking.

Imagine in my example that Person A was ranked 2nd in the interview and Person B was ranked first. Person A most likely - it depends on the actual score - gets the offer due to their superior scores in UMAT and ATAR.

ps - thanks for the "simplistic" comment. If anyone can look at these descriptions and say that then they are good descriptions!:)

I agree that your description is rather easy to follow and thus is a good example.

I think in your example though that if you scored second in the interview your z-score for the interview would be very high. Even if B scores a marginally higher z-score due to the first place rank vs second place rank, it wouldn't be enough to offset a 30 point difference in the other two domains because z-score is conceptually your scores as a deviation from the mean - this means that the first and second ranking students would have rather close z-scores considering the number of applicants numbers several thousand.

However, if we were to take an example where the raw interview scores were 100/100 for person B and 90/100 for person A, but as a result person B was ranked first and person A was ranked 100th, person B could then have a much higher z-score than person A to the point where it could account for that 30 point difference in the other two marks.
 
  • Like
Reactions: A1

A1

Rookie Doc
Moderator
I wrote this reply to a PM question, posting it here for any non-NSW/VIC applicants still contemplating doing both Unsw & Monash interviews.
____________

Last year Monash interview for interstate was on 27 Jan, I received the offer on 20th (exactly a week before intv date) so this year should be sometime next week.

Monash intv time is by online booking, therefore I suspect unless previously arranged it WON'T show you any slot for 25th and you are stuck with 24th.

Unsw sends an email giving you 10-12 time slots for the day (24th), you then reply indicating 4 slots in order of preference. They will try to arrange for you the highest pref possible.

Hopefully with this info you can plan in advance how to squeeze two interviews 900km apart into the same day.
 

shadowduke

Monash MD I
Hey A1, can you elaborate on this bit?

"unless previously arranged it WONT show you any slot for 25th"

How can we, therefore, arrange with Monash for the 25th?
 

A1

Rookie Doc
Moderator
Hey A1, can you elaborate on this bit?
"unless previously arranged it WONT show you any slot for 25th"
How can we, therefore, arrange with Monash for the 25th?

Last year Monash dates initially showed 27-28, I rang to ask how to ensure 27 as I wanted to reserve 28 for WSU. The lady replied 1st day 27th for Med, 2nd day 28th for other Health courses but (key info here) some Meds if they could not all fit to 27th.

Translate that to this year, 25th is generally not for Med but it can possibly accommodate some. I was thinking if you call Monash Med (not Admissions) and plead your case that 24th clashes with Unsw, maybe they will show compassion and give you 25th.
 

Registered  members with 100+ posts do not see Ads

Okay, so I'm slightly (read: incredibly) freaking out right now. So, my ATAR is 97.6 (IB 39) - was predicted at least 99.70 (IB 43) and then I screwed up exams. However, I might get it jumped up to 98.90 (IB 41) if remarks work out well. My UMAT is 96% percentile. So, UNSW is giving me a potential 5% (oh joy), so yeah, emotions are rampant at the moment.

My questions are:
  • would you say there's any point to the UNSW interview?
  • do you think there's a likelihood I can get into Monash?
 
Status
Not open for further replies.

Registered  members with 100+ posts do not see Ads

Top