Radiology 2012 applicants

[PokerDoc]

---

Members do not see this ad.

<- No interest in Miami. Interesting discourse nonetheless.

I'll be keeping my #1 to myself 😉

Amen.. I like Miami a lot, but as of now I've only gone on 3/19 scheduled interviews so far, so I really can't say I have a solid list worked out yet. I am really intrigued by some of the Boston and NYC programs I have coming up in a couple of weeks. I was just simply speaking to my own experiences rotating through medicine at Jackson that didn't seem to match up with what was being said on here, nothing more.

 

[2012mdc]

Amen.. I like Miami a lot, but as of now I've only gone on 3/19 scheduled interviews so far, so I really can't say I have a solid list worked out yet. I am really intrigued by some of the Boston and NYC programs I have coming up in a couple of weeks. I was just simply speaking to my own experiences rotating through medicine at Jackson that didn't seem to match up with what was being said on here, nothing more.

Do you plan on going on all 19? 16 interviews in 2 months seems like a lot but probably doable if you've done most of your prelim interviews.

I have done 2 and have 11 more scheduled with the possibility of 3 more. I have 4 in December and 7 in January but 5/7 are in the LA area

 

[PokerDoc]

Do you plan on going on all 19? 16 interviews in 2 months seems like a lot but probably doable if you've done most of your prelim interviews.

I have done 2 and have 11 more scheduled with the possibility of 3 more. I have 4 in December and 7 in January but 5/7 are in the LA area

No I want to cancel 2-4 more at least. I already cancelled 11. I dont think going on any more than 15 makes any sense, especially when I only have maybe 11 programs on my list I would actually be happy with.

I just haven't figured out which of the programs to cut out (prob atlantic in jersey, allegheny in pitt and st. vincent in mass are the easy ones)

 

[drizzt3117]

Why?

Doesn't statistically increase your odds of matching and if you can't find a program you like with 15 interviews then <shrug>. I've kept my list constant at 15, canceled 5 interviews so far.

 

[PokerDoc]

Doesn't statistically increase your odds of matching and if you can't find a program you like with 15 interviews then <shrug>. I've kept my list constant at 15, canceled 5 interviews so far.

This... plus if i go that far down my rank list, i dont think i wanna match at all lol

 

[DrBowtie]

I'm going on ~20 interviews because I want to evaluate places in person (even though it is only a brief interview day). My pre-interview and post interview ranking has changed pretty significantly for some programs.

I mean its fairly easy to cut other interviews if all 15 are in the top 25. I have some of those, but I'm looking at that middle chunk on my list to find some diamonds in the rough so to speak.

The only thing I'm spending is time and money. When you take into account the amount of money for all of your education, and all of the time you have spent in school, why not spend another 2 days and a couple hundred $ to see if its a place you fit?

 

[2012mdc]

This... plus if i go that far down my rank list, i dont think i wanna match at all lol

Same here.

I'm going on ~20 interviews because I want to evaluate places in person (even though it is only a brief interview day). My pre-interview and post interview ranking has changed pretty significantly for some programs.

I mean its fairly easy to cut other interviews if all 15 are in the top 25. I have some of those, but I'm looking at that middle chunk on my list to find some diamonds in the rough so to speak.

The only thing I'm spending is time and money. When you take into account the amount of money for all of your education, and all of the time you have spent in school, why not spend another 2 days and a couple hundred $ to see if its a place you fit?

Yeah for me it's not about the money, I don't think I could fit 20 in (if I had 20 invites, only have 13) because many of the programs that I wanted but didn't get only interview in Dec and Jan and some only in Jan. It would have been very difficult and exhausting to fit in 20 interviews in 2 months.

I actually want to drop a couple of interviews but I won't feel comfortable with only 11.

I've separated my list into tiers and I'm pretty sure there will only be movement within tiers and not between tiers.

 

[drizzt3117]

Interview fatigue is a serious concern, I'd be careful esp if some of your top choices are late. All my interviews are top 25 or in CA (uci/ucd) and I have a pretty clear top 4.

I'm going on ~20 interviews because I want to evaluate places in person (even though it is only a brief interview day). My pre-interview and post interview ranking has changed pretty significantly for some programs.

I mean its fairly easy to cut other interviews if all 15 are in the top 25. I have some of those, but I'm looking at that middle chunk on my list to find some diamonds in the rough so to speak.

The only thing I'm spending is time and money. When you take into account the amount of money for all of your education, and all of the time you have spent in school, why not spend another 2 days and a couple hundred $ to see if its a place you fit?

 

[DrBowtie]

Interview fatigue is a serious concern, I'd be careful esp if some of your top choices are late. All my interviews are top 25 or in CA (uci/ucd) and I have a pretty clear top 4.

You're right that fatigue is really the only downside, but its actually been pretty nice flow.

3 weeks of interviews (7) -> Thanksgiving + RSNA week break
3 weeks of interviews (6)-> Christmas + New Years break
4 weeks of interviews (7)-> Done

January is really the only time I think I'll really be struggling. Especially with my 2 prelim invites that I have in Feb

 

[badasshairday]

January is really the only time I think I'll really be struggling.

I think it is ok, because by that time everyone has a varying degree of fatigue. That's when you gotta put your game face on and march through it.

 

[2012mdc]

You're right that fatigue is really the only downside, but its actually been pretty nice flow.

3 weeks of interviews (7) -> Thanksgiving + RSNA week break
3 weeks of interviews (6)-> Christmas + New Years break
4 weeks of interviews (7)-> Done

January is really the only time I think I'll really be struggling. Especially with my 2 prelim invites that I have in Feb

Yeah that is reasonable. Having 7 interviews before December makes it doable

I'm in a similar boat to drizz (just a heavier Cali slant) and it's possible I could have as many as 10 interviews in January including 2 LA to Northeast roundtrips. The only saving grace is that 8/10 would be in Cali and 7 I could drive to.

 

[drizzt3117]

.

 

[2012mdc]

lol...not to be an a-hole, but then why bother even going on the interview and ranking the program?

http://www.nrmp.org/data/chartingoutcomes2011.pdf

Gentlemen, if you would go to page 51 (47 in document) please...if I am interpreting this right, the risk of not matching with 15 interviews is about 3% whereas with 20 it seems less than 1%. For indies, the difference is even greater...maybe 7% versus 4%.

I know these do not seem like big numbers, but a 3% difference means that 1 in 33 peeps here will not match anywhere despite going on and ranking 15 interviews. There are probably 33 people posting in this forum itself. Does this matter or do your bottom 5 programs suck enough that you'd rather be nowhere?

The biggest thing is that you don't know the overall strength of the applicants or the composition of the rank lists.

Someone who is a very strong applicant and has non top 25 programs at the end of their ROL of 15 probably has less than a 1% chance of going unmatched.

Someone who is good applicant and has a ROL of only top 25 and competitive non top 25 in their ROL of 15 probably has a much higher than 3% chance of going unmatched

 

[drizzt3117]

.

 

[bigdoc12]

I may be way off here, but I think that if you have >10 interviews, your chances of matching may actually increase by cancelling interviews that will likely by at the bottom of your ROL - thus allowing other people to pickup these cancelled interviews, possibly match at said places, and open up more spots at the top of your ROL....purely theoretical and depends on program's strength, obviously

 

[drizzt3117]

..

 

[2012mdc]

Can't argue with that. That said, all else being equal, if you have 20 or more ranked, it is more likely that you are very competitive at the bottom of your list than if you had 15 or less.

Again it depends on the composition of one's ROL

For me personally if I had 20 interviews instead of 13, the program that's last on my ROL of 13 would be last on my ROL of 20.

It's just impossible to extrapolate the difference between 1 and 3% to the individual level.

Do you even know the p-value? It may not even be statistically significant, let alone clinicaly.

 

[DrBowtie]

Again, can't argue with that. But you may have more safeties in the list of 20, even if the one in last place is the same. I guess my point was to make sure that I was not crazy and that more than any other factor (including Step 1), the number of contiguous ranks is most correlated to matching somewhere. It is the only category that hugs the 100% line so closely in the probability curves so I guess it makes sense.

Everyone has their own risk tolerance. Everyone is entitled to interview at however many they want without criticism.

 

[drizzt3117]

Ummmm, yeah...I am saying the people with 260+ were not ranking only top 25 programs. That seems to include both myself and other people in this thread. And quite frankly, if you read pg 54 (calculated based on USS 2009-2011), it very clearly shows that there is a real risk even in the 260+ range. At ease, man...didn't mean to rile you up. I just wanted to make sure I was reading the data right.

One of my fellows on one of my aways told me he interviewed at 50 programs. If you're that risk adverse, you can always go that route.

I don't think it's clear at all regarding step 1. The graph is plotted based on data with intervals of 10%. You have no idea whether its 1% or 3%.

Furthermore, the graph with % contiguous ranks doesn't include step 1 and vice versa. Therefore, although we don't know for sure, I would bet money that the number of people who ranked 15 programs and had step 1 > 260 would be rare by anyone's standards. If we made a graph of applicants with step 1> 260,i bet it'd be hugging 100% by 10 programs ranked if not earlier. Our dean says that in the last 10 years, our internal #s have shown that of applicants with step 1 > 250 (a lower bar, and a group of about 40 people at my institution, historically we've put about 11 ppl into rads on average, but not this year) not one single person has failed to match. He recommended ranking 12 programs with my application: I figured I'd play it safe.

Anyways, I don't think there should be an arbitrary number that one sets as the # of programs they should rank. The programs at which I chose to interview at were all ones I'm interested in going to. This isn't necessarily ranking dependent. I chose to cancel interviews at top 15 programs, for various reasons, it's not simply that im dropping the lowest ranked programs on my list. I also haven't dropped any programs I'm legitimately interested in visiting.

 

[2012mdc]

Again, can't argue with that. But you may have more safeties in the list of 20, even if the one in last place is the same. I guess my point was to make sure that I was not crazy and that more than any other factor (including Step 1), the number of contiguous ranks is most correlated to matching somewhere. It is the only category that hugs the 100% line so closely in the probability curves so I guess it makes sense.

A lot of safeties won't even interview people they assume are using them as safeties (a major reason I have 13 invites even though I applied to 45 programs with strong stats)

There is no proof that the difference is statistically significant.

Also number of contigous ranks is the most dependent on other factors.

There is a huge difference between an applicant who applied to 100 programs and only got 15 interviews and ranked all of them and the applicant who who applied to 30 programs and got 20 top 25 invites and ranked 15 of them.

One probably would benefit from having a longer ROL and the other probably wouldn't. They shouldn't be treated the same

 

[drizzt3117]

1. Again, don't mean to rile you up and I have no real vested interest here but...It is obvious that the risk is not even close to 0%, as you suggested, and certainly not laughable. Not knowing exactly what this risk is is even more reason to be conservative. Also, if you just visually integrate the area above the curve from 260 to 280, the risk is easily at 3%, possibly even 5.

2. It is not arbitrary if it is based on a probability curve calculated by the NRMP based on years of applicant data.

Anyway, that's it for me...I think it's pretty obvious by now. Cheers fellas and best wishes for the match 😉

Let's not let facts get in the way of a good argument. If you can't read those numbers, it's 95% = 238, 99.2% = 260.

 

[PokerDoc]

Don't forget that objective data can have subjective causes.. say those 3% who didn't match were sociopaths or the most awkward people on the planet? No board score will help you if you're not someone people want to be anywhere near.

 

[badasshairday]

If you are able to rank ~12 programs and are not a sociopath, you should be good. Those of us doing 12+ are doing it for insurance or to check out a couple of more places we are interested in.

 

[drizzt3117]

I am responding mainly because I applaud your effort and I also realize that saying anything at all was a bit insensitive of me given that people's careers are on the line. But you kinda have to add every percent from 260 onward, not just at that point. That would be 0.8% + 0.7% + 0.7% etc. until the curve ends at around 270. Hence: "if you just visually integrate the area above the curve from 260 to 280, the risk is easily at 3%, possibly even 5." Just imagine flipping the curve upside down and then getting the AUC from the line you drew onward.

Since this kinda thing is funny to me, what would be cool is to get the actual equation of the curve and use calculus using integral limits 260 to 280 and get a definitive answer.

No you do not. The Y axis is % chance to match.

You only have one score, which means you have one percentage to match on the curve. If you have a 238, you have a 95% chance to match based on past data. If you have a 260, you have a 99.2% chance to match.

Just think about what you said for a second. What you're saying is that if you had a 238, you'd add 5% for that, 4.5% for 239, 4% for 238, etc. Where as we know that the national % chance to match for US applicants is 96% (which makes sense because the national average is 96% (883/919)

 

[drizzt3117]

I could be wrong...but my understanding is that these curves were created/averaged such that for each point, say 260, 0.8% did not match. For 261, 0.7% did not match. These curves are not so much predictive as reflective of past data. This is the only way that they could set up the curve using past match data, as far as I know. That said, I could very well be wrong here, which is why I was interested in everyone's input.

You're 100% wrong. 96% of all US applicants last year matched, yet you're trying to say that you have a 95% chance to match if you have a 260? lol

But the fact remains, however you analyze the data, that the risk is closer to 1% at best even for a score as rare as 260, which contradicts what you originally so adamantly stated.

Again, misinterpreting what I said. I said that the chance is astronomically low given 1) the score of 260 AND 2) 15 contiguous ranks. In this situation, it's summative, so the percentage is higher than either individual percentage. My guess is that it'd be somewhere between 99.6-99.8%, so about 1 in 250 applicants with these stats wouldn't match, or 1 in 500, meaning that one person with this profile would fail to match in about 4-8 years. That sounds rare to me.

The whole discussion is just ridiculous. Look at the average stats of the unmatched applicants. Mean contiguous ranks = 3.6, mean step 1 = 211, % AOA = 5.6%. We haven't even taken grades into account in this equation. It's patently obvious what kind of applicants aren't matching. Those with below average stats and not a lot of interviews.

 

[2012mdc]

Obviously that is not true when all 15 programs are top 25 programs.

do you have any facts to back that up?

there are top 25 programs that aren't hyper competitive bc of location

 

[drizzt3117]

.

 

[drizzt3117]

.

 

[drizzt3117]

.

 

[drizzt3117]

But how does that make sense...if the probability of matching with a 238 is about 96% and about half of the matched applicants have less than that score...how can 96% overall have the same %match as someone with a 238?

You're making the incorrect assumption that 96% of the matched applicants have less than that score, it's not a median score, it's a mean score. Again, 238 is about a 95% chance to match based on the historical data. About 200 applicants have < 230.

 

[cowme]

I can speak from personal experience that if the top of your ROL is top 15 programs, it is pretty easy to slide way down your list. No, it doesn't mean I'm a sociopath, but rather these programs are in high, high demand, and tougher to match at than you believe. IMO, you'd be taking an enormous risk by only ranking high demand programs.

 

[drizzt3117]

I can speak from personal experience that if the top of your ROL is top 15 programs, it is pretty easy to slide way down your list. No, it doesn't mean I'm a sociopath, but rather these programs are in high, high demand, and tougher to match at than you believe. IMO, you'd be taking an enormous risk by only ranking high demand programs.

Another factor to take into account is the ~20% decrease in apps for the season compared to past seasons, with either the same or increased scores for the top applicants. This is even more likely to create a situation in which the top programs are ranking the same people, as has been discussed on AM.

 

[drizzt3117]

How does that make sense? So someone with a 238 has a less chance of matching than the entire applicant pool overall, including those who didn't match?

Sorry, the average for matched US seniors is 240, not all US seniors. The difference there is simply based on the fact that the 96% is from last year's data where as the graph shows data from 2009-2011.

 

[2012mdc]

That should not change the order of their list. If the person does not want to go there or wants to go to a program higher on the applicant ROL, then the program's ROL would simply go to the next person.

It doesn't work that way.

If I interview somewhere and I am polite/sociable/interview well but someone in the program overhears me telling another applicant that I consider this program a backup I will most definitely move down on their ROL and may come off it completely.

If a program doesn't think you're truly interested in the program they won't rank you highly or rank you at all.

 

[drizzt3117]

...

 

[drizzt3117]

My problem with this is that the probability curve for contiguous ranks does not support this.

The probability curve is based on applicants with an average step 1 of ~237 or so whereas if you're talking about applicants getting interviews from the bottom 25 programs, likely have a lower score than that. Again, many factors influence if you're going to match, including but not limited to grades, step 1, and how many places you interview at and match.

 

[drizzt3117]

Dude...of COURSE you're right. We're talking about when it is not totally obvious that you think the program sucks...when the PD is trying to read your mind, gauging your likelihood of ranking them highly based solely on your application despite a perferct interview performance.

The whole concept of yield protect and trying to predict other people's decisions would be governed by game theory, except that game theory and the concept of a prisoner's dilemma implies no communication between the parties. Obviously, the parties are able to communicate, so the situation you'll have is one in which the PDs are going to pressure the applicants to tell them what they really think about the program in many cases.

 

[rem6775]

I'm having to scroll through way too many posts to make sure I'm not missing anything important.

 

[drizzt3117]

..

 

[drizzt3117]

I'm having to scroll through way too many posts to make sure I'm not missing anything important.

I think a lot of this discussion is irrelevant but I do think that the point we're discussing now about how programs rank applicants and how applicants need to communicate is quite important. I predict it's going to be a major issue in the next two months for me. I have a top 4 that is pretty much set and won't change, so figuring out how to go forward from here not only in terms of making a decision once I've interviewed at all four programs and communicating with them is going to be hard.

 

[MedManU]

i'm having to scroll through way too many posts to make sure i'm not missing anything important.

+1

 

[DrBowtie]

Anyone have any extra info on this? Implications on residency program at DMC?
http://www.auntminnie.com/forum/tm.aspx?m=327911

 

[drizzt3117]

This is offset by the fact that as a group, the average scores of their competition is similarly low.

This would be true except for people applying broadly like you're advocating.

 

[drizzt3117]

Well, they are just not mind readers and PDs admit that they are routinely fooled by applicants. It happens so often that, according to one NRMP report, 80-90% of PDs do not believe applicants' supposed preferences.

source? Also, what PDs?

Also, just to play the devil's advocate to the point of absurdity 😛, based on our psych lectures in med school, sociopaths tend to interview better and are less likely to drop down their ROL! 😱

🙄

 

[badasshairday]

Neuro-IR is cool.

 

[2012mdc]

I'm starting to believe we're getting trolled. The poster is just disagreeing with anything that is being said regardless of the topic

 

[drizzt3117]

Neuro-IR is cool.

It's pretty awesome. I've kinda gotten away from it lately though, I think int onc has a bit brighter future and as noted is significantly faster to get into. All of my research is in neuro-IR, though.

 

[2012mdc]

It's pretty awesome. I've kinda gotten away from it lately though, I think int onc has a bit brighter future and as noted is significantly faster to get into. All of my research is in neuro-IR, though.

Int Onc is what first drew me to IR

 

[jahbrony]

I'm starting to believe we're getting trolled. The poster is just disagreeing with anything that is being said regardless of the topic

+1. If you recall, there have already been a few people who trolled this thread. It wouldn't surprise me if some of the ones posting were reincarnations of previous trolls.

 

[SGuttenberg]

Earlier someone posted about no 260+ step 1 people not matching, well I think if it happens it's going to be me. I got >260 but have only 3 top 25s (applied to 12 of them) and tons of rejections. Out of sixty some programs applied to I only had 20 interviews.

Step scores are a screen only. I'm sure there are some 240s step 1s at mgh and ucsf. I don't think anybody will care if you have a 268 vs. 248 if you're a major stud otherwise.