USMLE STEP 1 BIOSTATS: Exam REVIEW Q's
Вставка
- Опубліковано 22 лис 2024
- ESSENTIAL MATERIALS
www.amazon.com....
Disclaimer: As an Amazon Associate I earn from qualifying purchases. There is no additional charge to you.
Are you ready for the STEP exam? This video reviews question that the makers of the USMLE are expecting you to be able to answer at the minimum. The concepts are same from previous videos, but the questions are asked in a manner where you must have mastery of the concepts. Hope this helps...
Randy Neil, MD
A good point brought up by a viewer was the time it is taking to complete these problems. Our goals is to review as many types of questions, so that we don't fear however they may test us (on real exam). The writer's of the USMLE STEP exams want to make sure you have the CONCEPTS down. It may take me several minutes to work (teach) one problem, but it's so the time to answer becomes less and less each time we see the next 'similar' problem... Our goal is to keep the BIOSTATS problems on the exam 1 minute or less, so we can take the extra time on the other topics. BIOSTATS needs to be our GUARANTEED POINTS on the exam. That is our goal, our purpose here... Hope this helps. Keep moving forward.
~Dr. Neil
Thank you for these amazing videos .in this video at ua-cam.com/video/xDgDpEDs1vg/v-deo.html I am confused that you are going in wrong direction and coincidentally answer remain the same in either direction. you are going in right direction instead of going down. this is my confusion I might be wrong but it confuses me. Thank you.
thank you ...ive developed SO MUCH anxiety over this. seem like i study study study...get anxious about relaxing and study...mind races and i think about everything I DON'T KNOW...then I sit and listen ...feel I know it...only to forget in two days. I'm a basket case over this.
@@morganlucas9062 Keep reinforcing it. It's like memorizing a Birthday...eventually you just KNOW it.
Hello Dr Randy. Is it possible to make a video in androgen insensitivity and other gender disorers. Dont know why I am weak in that too. Thanks. 🙏
الله يسعدك و يوفقك
I wrote in my native language ( Arabic ) trying to express my gratitude for you ✨
It means * God bless you and grant you success*🌟💐
Thank you Ethar. I appreciate the kind words.
please keep reviewing more of these..really help...one of the mnemonics that I learned from dirty medicine for sensitivity and specificity went something like this...'' if you're sensitive you're always going to go DOWN in life. if you're specific. you're always going to go UP in life''..hope it helps anyone to keep things straight
Waleed, Thank you for sharing!
Thank you Waleed :p
No joke... watched the biostat videos yesterday and just destroyed my stats blocks on combank. In my head I had I kept saying "this is all smoke and mirrors". You're the real MVP Dr. Neil!
Sharon, Glad it was helpful... I appreciate the feedback!!
Hi Dr Randy! Thank you for your great videos they have helped me greatly!
I was a bit confused @12:28 so I had to look it up. When you decrease power:
1. FP (Type I Errors) Remain Unchanged [the rate of FP is determined by the significance level (alpha) you set for your test, typically 0.05; decreasing power does not directly affect the rate of FP, as this is controlled by your alpha level].
2. False Negatives (Type II Errors) Increase: Lower power means a higher chance of failing to detect a true effect, leading to more FN. This occurs because the test is less sensitive to detecting differences when they actually exist.
Please correct me if I'm wrong. Thanks again for all you do! You rock!
i agree , explains the confusion . the alpha error is preset so would not be influenced by the changes you make to the sample its going to be 0.05 no matter you increase or decrease the sample size.
Your videos are hands down the BEST videos for Biostats. I really appreciate you taking the time to show even the little steps, cause that really makes things stick! I always do your videos before my exams and feel super confident! Thank you
Green, Thank you for the comment. Glad they are helpful... One STEP closer!!
there is no words to describe this videos. A true treasure. It will be awesone if one day you decide to have a platform , where you can teach all Biostat.. because as IMG back in my country we did not receive the asignature perse. This a must for every IMG
Reynaldo, Thank you for the kind words. Glad it was helpful.
I love, I love your techniques and your vibes that you emit with your positivity and uplift us for our studies. Really missed you Randy! Great come back. Wish you all the best of luck and success wherever you go 🙏🏻💖💖💖
Can't wait for your behavioral science videos 🙈❤❤❤
Thank you for feedback... I'll do my best to put something together in those other subjects. As long as there is a need, It's a fun process teaching/helping... Keep us updated on your progress...
About question number 1: shouldn't the correct answer be B instead of D?
They asked about the False negaitive in order for us to be able to calculate Sensitivity (TP/TP+FN) and we already have the TP (90)
I thought the answer choice B actually describes what we are looking for (FN)
Am I getting this wrong?
i think the same way as you
X3, I was thinking B would be the FN to calculate the sensitivity.
Hey B is basically describing false positive not false negative . And X as directed on the table is for false negatives . So B isnt correct
I was thinking the same
actually the false negatives are detected by the gold standard test (in this case, biopsy). False negatives do have the disease in reality but the test makes it look like they don’t. As we don’t know if the other 200 patients were true negatives or the test made them look like that. doing choice E would find out FN. hope this help
At 18:23 u added who took the drug and who didnt,u should have said who took the drug and died +who didnt die after taking the drug. As the values were same it didnt affect the answer but if the values wern't same the answer would have gotten different.
Thanks Doc for the effort you made to make our life and test easy !
Maryam, Thank you for taking time to comment...
Thank you so very much, step 1 in a month and you are a life savior..totally going to subscribe.
Thank you Bini
Thank you. I can't tell you how much I appreciate your explanations.
Thank you DocA for taking time to comment.
Another excellent video Randy! I would encourage you to change the titles of your stats videos to Step 1 & 3, since I keep hearing that day 1 of the Step 3 exam is very biostat heavy. Fortunately, I found your videos and will be prepared :)
David, Thanks for the comment and TIP... Thanks for taking the time.
Hey Dr. Neil when you are doing the calculations at 18:48, you draw an arrow to the right however shouldn't it be an arrow downwards? You're calculating individuals taking X vs those who do not so you want to do the column of + drug X first over the column of not taking drug x?
Oh my god!! Thank you so much for doing this video.Thank you so much for listening to our requests and uploading them. I am pretty confident now going into my exam on monday.
Zlatan, Confidence is key... Appreciate the feedback and comment. Kick BU## on exam and then help next guy in line... Keep moving forward.
@@RandyNeilMD Absolutely sir. Thank you.
Thank you very much for your nice videos,
I think question in 14:20
The Table is wrong : ( I think following order is the right)
As Intervention or test in vertical lines or side of table
While Disease or Result e.g Die or alive on top table boxes
Thanks again for your great efforts .
Let me know if that's right or not.
I thought the same
not sure if I should go with this or if my concept is wrong
He messed up the correct orientation. Drug x should go left to right row and cancer status should go top to bottom -vertical column, but he kept the numbers in same boxes so the math came out correct
Question 1 - first they picked 300 males , then checked their serum PSA , 100 were found to have > 5 . They let the remaining 200 males go and isolated these 100 for biopsy to confirm Dx. 90/100 had positive biopsy findings for cancer ( TP ) . 5/100 had prostatitis findings ( FN ) . B seems right . D is wrong because only those men underwent biopsy whose PSA was GREATER than 5 ( not equal or less )
I am addicted of your out standing knowledge of Epi/BioS 😉😍😍😍👏👏👏🌺🌺🌺🌺
Humbled.
Thank you so much. God bless you.❤
You are most welcome Tahera... :)
FOr drug and cancer question, It would be easy and a good way to always keep reality like disease or disease outcome like death in top and exposure or test in side. There is less chance of confusions and mistakes.
Sandeep, Yes, agree. Consistency is key. Thank you for the comment.
Always a pleasure seeing your videos. You are great at this. Hope to see you publish a textbook.
Keron, If we put BIOSTATS in a textbook, then we could cure INSOMNIA... Something about old school teaching that mimics how it might feel on the exam is what I feel might help us learn. I appreciate your confidence in me and thank you sincerely for the comment.
Hi Dr! Please I need more explanation on the second to last question you answered.
Hi Dr. Neil, in question 2, why are we using the erythromycin numbers if the question specifically only asked about azithromycin?
Because in this question the erythromycin represent the control group and the azthromycin represent the event group.
Dr. Randy could you please explain why at 12:46 the p value decreases with decreased power/no. of participants in the study? Wouldnt that imply the results become more clinically significant with less participants in the study? My understanding was greater participants=greater power= more chances of your results being statistically significant =lower p- value= less chances of type 1 error?
Thanks for the video. I couldn't understand the change in alpha that you mentioned at 12:25. I'd appreciate it if you could explain this part
We "assume" ALPHA would rise if we make changes that make a study worse, per se... like decreasing the POWER (number of people). Decreasing the Number of people in study would make for increased chance of error...thus increase in ALPHA. Does this help?
@@RandyNeilMD I'm sorry, but I don't understand the relation between the 'alpha' error and number of people in a study. Also, I assumed that 'alpha' and a type 1 error referred to the same thing, i.e. a false positive error. In your statement above, you refer to a direct relation between alpha and a type 1 error. However, in the video, you have mentioned an inverse relation between alpha and a type 1 error.
@@RandyNeilMD I thought the type 2 error, and not the type 1 error would be increased by decreasing the number of people in a study.
Your correct, type 2 error (beta decreases, not alpha) if you increase the sample size, effect of size, or precision
Yeah just got a similar question in an NBME. ↓ing Sample Size decreases Power & Increases chance of Beta Error.
Alpha value will always remain same unless question stem mentions about changing p-value (which corresponds to alpha value).
Same thing mentioned in Uworld Biostat Qbank too.
thanks. please do a few different types of hardy weinberg questions
Sandeep, Thanks for the feedback. If I can find some H-W Examples, I'll make sure to post... The hardest part is getting good questions.
@@RandyNeilMD
Thanks for the reply. Looking forward
Q1) know how to define each component in the formula of sensitivity & specificity ; know the formula.
Q2) 2:47 => ARR = | Ec-Et| where event means incidence
Q3) 5:11 =>
Confidence interval must not contain one else it’s a useless study (not significant)
If RR is 1 means no association
RR less than one means protective
RR more than one means risk association is present
7:25 q4) calculate specificity sensitivity from pre-made 2x2 tables :))))🎉🎉
Q5) how to interpret any study results?
Step 1 : Is study statistically significant? P values must be less than or 0.05 & CI must not include
Q6)10.02 : must know how to draw null hypothesis box & must know that n is directly proportional to power of study
Q7) 13:44 relative risk
Must know how to draw 2x2 table
Plug & chug
Q8) 20.19 covariants or interpretation of data
Whenever you see CI, chuck out the garbage data Ie which includes one.
If odd ratio or relative risk is more than one: risk Association present
If equal to one : no association
If less than one : protective Association present
Q9) 23:19 => comparing & interpretation of data
If there are two groups :
Which group showed more improvement when taking x intervention as compared to placebo? (Here: physician vs pt)
Step1 : check if the data is significant (p
thank you Dr. Neil
Most welcome.
I love this ...however...I'm a little slow and it goes rather quick. I know I need to catch up because time goes swiftly on the real exam...
Yeh, I wanted to increase the pace for a better review...
Thanks for your great efforts.
You are welcome.
awesome...thanks Dr Neil
Rebel, Thank you for the comment!
So helpful, thank you very much, Dr.Neil!
Dekuang, Glad it was helpful. Thanks for the comment.
Thank you so much doctor Randy for the great effort :) you've made biostats way more interesting .
in the minute 12:40 why would the chance of type 1 error decrease if alpha increases ? isn't it the opposite ?
same thing I am wondering and confused about
Me too didnt get that part:(
Alpha (type 1 error) is NOT reliant on population size or accuracy of the test (in contrast to power which is). Alpha value is set by investigators before the experiment and is independent of the design of the study. Usually alpha is judged against p value to see what the chances were that they made a type 1 error but again, it does not change
love the way you teach , nice techniques
Thank you... I appreciate the feedback.
Dr. Neil, can you explain the confidence interval and sampling size?
so informative! I have a question @26:33 you wrote the equation control - treatment & calculated treatment - control?
i had the same question.... a year later
thanx a lot Dr.Neil
Hossam, Glad it was helpful... Appreciate the comment.
Thank you so much for such very helpful review for Biostatistics for step 1. Dr. Neil, is there any chance to upload videos for Step 2 CK Biostatistics? I found those questions regarding external validity "generalizability", Internal validity, reliability, reproducibility, selective survival bias, confounding and effect modification are confusing in Step 2 CK Biostatistics.
Shaima, You are putting me to work... Let me see what I have for those topics. Thank you for the comment...
@@RandyNeilMD Greatly appreciated!
@@amirdada5700 yes these step 2 ck internal validity and selective survival bias kind of problems are very annoyying.. Please hep me Dr randy Neil
Thank you so much for these great videos ! Can I ask why the first question's answer isn't B ? Because I thought my reality is all men with SPA concentrations GREATER than 5 ng\ml.
In 1st question 100 is blood test results positive so equals True positive plus False positive (not the total no. Of diseased individuals) please correct me if i am wrong @RandyNeilMD
I have the same doubt
I would really want you to go over the first question, again. The correct answer is B not D! Thanks a lot for all the efforts :)
Hey B is basically describing false positive not false negative . And X as directed on the table is for false negatives . So B isnt correct
Hello,
There's this confusion. At @26:15, wouldn't NNT be 12 because you took the percentages already mentioned in the data as values for ARR which were a/b but ARR in the formula is a/a+b so that would change the answer. Please explain if there is any other way to solve this, which I didn't understand.TIA.
Thank you so much!! really helpful video and your technique and explanations are also good
Saket, Thank you again for the comment.
Thank you🙏🙏
You are welcome.
You are a gem! Thank you so much for these videos.
Ankita, Thank you for commenting. Glad video was helpful.
I don't know to say
Thanks dr randy neil
Thank You so so much for these......I learn so much from your videos and have been binge watching all your videos and questions and learning a lot....please keep them coming with different topics.....is it possible to do more NBME style questions from CBSSA?
Thank you for the comment. I will definitely try to keep some variety... I', trying to get all the common stuff first... If you have any suggestions, I can always see if I have the material.
I am a little confused on 1st Q if u take reality(cancer) on top and test(biopsy) on left then ppl with positive biopsy and CA also known as TP will be 90 as far as im with you but in sensitivity we take TP on top and TP+FN in bottom so False negative are the ones who have prostate Ca in reality but biopsy says they dont this makes it FN which is required for sensitivity so isn't answer choice B correct?
also D talks about PSA equal or below 5 but we didn't biopsy any below 5 all 100 biopsies performed were greater then 5.
Your response will be appreciated
yea unfortunately Ive seen some inconsistencies on how he calculates too...its confusing from what we see in the book....but atleast he is still great with concept but for questions?? I suggest Uworld after watching most his videos
@@nevmitch2 Yeah i am doing UW as well. can't thank him enough for lectures. I think minor errors are okay since his lectures so good i am able to do the questions with such ease.
dr.neil at 12:55 why did the type 1 error decrease ?
I;m hoping to figure this one out as well, it was decreased then increased.
It’s incorrect, sample size only affects beta error (not alpha error). Please do keep in mind he not a biostat expert. He knows enough to answer usmle type biostat questions and occasionally makes mistakes but most of it is error free
@@deathkissgoodbye yea i ended up figuring it out thank you.
Bec in type 2 error your study is saying there is a difference but there is no difference in reality. That error will decrease if your sample size increases to where you can confidently say there is difference if you have a lot of people in your study and will less likely to commit a type 2 error (decrease beta error) and increase the power of your study saying there is really a difference present in study
You can’t say that for type 1 (alpha) error bec from the beggining your study is saying there is no difference, so increasing the number of people in your study is not going to change that status that there is no difference bec having more people will help you see a difference if there is one to see in the study
alpha is the probability of making a type 1 error so that would increase chance of type 1 error with decreased sample size? can someone please confirm cuz in the video Dr.Neil has two opposite arrows for alpha!
Hi Sir Plz make a lecture on Confidence Interval and Mean Median Mode, Also do some problems of Incidence and Prevalence, Thanks in Advance!
I love your videos
I’m confused about your explanation for alpha or type I error in the problem at 12:40. Seems like you have it backwards?
Confidence Interval must not contain 0 or 1??
Dr Neil can you please do videos on other maths or equation related topic, like half life, or kidney or respiratory equations..??
I'll have to see what I have...
Thanks so much I found it helpful.
Simret, You are most welcome.
Thanks alot
Noon, You are most welcome.
Which video is the original you refer to throughout this playlist?
"Part one".... the "most watched" video...
Thank you so much
Ankara, You are most welcome.
You're the BEST! Thanks :)
Sarah, Thank you for the comment. Glad it was helpful.
@@RandyNeilMD T tests / Chi square would be very helpful as well!
Amazing video sir, I really appreciate your work. I had a doubt regarding 2nd question where u used erythromycin data whereas shouldn’t we do within the azithromycin group for example: 45-5 divided by 50( ARR of vomiting in azithro group
Foundation trainee in UK?
Great lectures 👍
Glad it was helpful.
Hi :D I´ve been watching your videos and found it very helpful. I am taking my test on Friday. Anyway, I was wondering about the question you did at min 18 about the relative risk and think you made a mistake when you were filling the formula with the numbers. However, you got the same RR because the numbers were the same. I got that from your other videos
Yeah exactly he made mistake in making ideal 2×2 table
Anything is possible...
Thanks Dr.Randy you are amazing ...I was wondering why did you change the rule from having the reality ( disease) test (investigation or drug etc). I found it bit confusing to change the rules to get the table right, especially the exam of the Drug x and prostate cancer I try to solve it the old way i learned from your video and i got the result right, my question is shall i consider putting on top ( reality) whatever comes first then i try to solve it or shall i stick the old rule you mentioned in the previous videos? best regard Sami.
Great question... I was worried that might confuse people. I should have been more consistent...but I knew that as long as I know what ever comes first goes on top and whatever is second goes on bottom, then it really doesn't matter which AXIS you put reality and test for these problems... For the sensitivity, specificity, null hypothesis problems IT DOES.... BUT for the Relative Risk, ODDs Ratio problems...Its all about making sure you "read it in order"...and AXIS doesn't matter. Does this make sense...? If Not I can try and clarify with a problem that would show "BOTH WAYS".
Hope this helps...
@@RandyNeilMD I understand now what you mean and it make sense now you put it this way thanks?
also a an example would be a nice gesture too if thats not an issue.. I love your videos and i have learned so much from you so I really cant thank you enough
@@arsenaldream2156 can you kidnly explain how you solved it with the old method.i just CANT get my head around this one
In 4 question 4 isn't the alpha remains constant no matter what sample size is??
Could you please explain that part about how the alpha error doesn’t change with the number of subjects decreasing from n=1000 to 500?
Thank u soooo much..
Your Welcome.
Again, I thought that first aid states that we dont want the CI to include 0 and we dont want the RR or the OR to include 1
if alpha is probability of making type 1 error, how does increasing it reduce the chance of type 1 error??
also, in RR of drug x on death you did died/not died instead of took/not took. luckly numbers were same in each way so result was correct
I had the same idea and the numbers are the same but if they are different numbers we will get different values!
Its a great video but on 18.09 just little confusion you are taking which ppl are dying table the ppl who took the drug that is on second table but since value are same so i think it Does not effects the answer. but thank you
Yes this puzzled me
Hi Dr. Neil, How do I locate your notes from the first lecture?. Thank you very much for all you do
Either send email to randy.neil8@gmail.com or go to CoastalMDlive.com... Hope this helps.
@@RandyNeilMD Preciate you.
Can anyone please help me understand how confidence interval crossed 1? I didn't understand the concept
Respect from Pakistan ❤
Glad it was helpful.
For the fifth question, how do you know whether it is clinically significant or not?
i believe he using the logic if not a significant difference in the 2 values (days) for the patient
@@nevmitch2 It isn't even a 2 day decrease though, which you could argue is significant vs the normal 6.7 day length. The decrease was from 6.7 days down to 6.4 which is only a decrease in 0.3 days. Nobody is going to want to pay and take a medicine that only reduces duration by 0.3 of a day
i did not understand your that question explaining increase chance of type 1 and 2 errors.
am sorry sir but the power of you memorizing the questions increase the power and decrease type 2 error I knew you memorized it in the hypothesis question
I like you but this video you kinda confusing
One thousand one nights of stats! :-)
I'm expecting a PERFECT score in BIOSTATS for you...
@@RandyNeilMD thanks for the support!
is there anyway to print out these questions?
This specific video were questions free for everyone on the USMLE website...
if they had a PSA of 5 or less then they never got a biopsy according to the vignette so D is wrong. These Pts would have never had the biopsy.
cool
👍🏻
Hi Dr. thankyou for making biostats easy for us ... though am bit confuse at 22.06 why we are rejecting the lung infilterate point coz confidence interval crosses 1 ...plz can you explain i didn't get that point ..others are also above 1.. anyone for help..thanks in advance , am confused in CI thing
Taranum, I like to eliminate choices in certain questions when CI crosses 1 because it is essentially telling me that there was no difference (Benefit or worsening) between the two groups (at some point both groups were equal)... In my experience it can be a quick way to eliminate a distractor. Not saying it's fullproof system but has worked for me.
@@RandyNeilMD thanks
❤️❤️😍😍👏👏💐💐
👍🏻
That`s great, but we have to remember that when rushing to answer the question, it is more likely to make a mistake as you spent nearly 7 min on a question!
Ann, Agreed, we have to pace ourselves (even when studying)... roughly 1.2 - 1.5 minutes per question. I'm hoping we can reduced the biostats questions down to a minute or less... leaving us extra seconds for other topics. The hope with the videos is to see as many scenarios, to where we can then just say "I've seen this before, I know the concept"... ignore the distractions and feel good about our answer.
I always have the intent to "go faster", but the "teacher" in me...pauses, and wants to explain. When actually studying for biostats, I recommend having the common 30 Biostats problems written out and can flip through the finished (written out) problems and just work it 'mentally'.
Hope this helps...
🙏🙏💐💐💐💐
👍🏻
i know this guy means well but Im not seeing any consistency in the calculations of some of these problems.....videos are still good for concept tho
Neville, Try the biostats videos from Texas from 2015... they start from scratch. This is just one way of doing it...
The ruddy impulse fortunately practise because icebreaker pivotally smash round a faint fair size. hushed, remarkable biology
I’m confused about your explanation for alpha or type I error in the problem at 12:40. Seems like you have it backwards?