Before beginning the lectures for today, please fill out one of the surveys below.
If your last name begins with letters A-L, take survey 1.
If your last name begins with letters M-Z, take survey 2.
Remember to record your response to the survey in your private notes. You will refer back to this response later during the lecture.
Then, watch both parts of the lecture by clicking here, or go to Blackboard, then navigate to: Content–Guest Lectures. Here you can also find “Closing in on Cancer,”the article Prof. Bayaz-Ozturk mentioned in her lecture.
After watching both parts, please make comments according to the instructions below. These comments are due before the following class period (i.e. before 4:00pm on Tuesday 11/2) and will only count towards your grade if made before that time.
The first one is your own individual thoughts and response to the lecture. This post should be at least 75 words long. You should interact with main ideas, although what you say is up to you. You may want to criticize an idea, agree with something, or offer further insight or thoughts about some topic discussed by our guest. Or, you may want to pose a question for others about the lecture. I expect the post to interact with the content of the actual lecture, and not just the day’s topic or other thoughts you may be having.
The second post is an interactive comment, posted as a response to another student’s comment. There is no length requirement on this post.
Prof. Bayaz-Oturk talks about how probability deals with axiomatic theories and how we take these theories and we understand them as rules when it comes to solving problems. Even though these rules may or may not be true, we believe them to be. I find it fascinating how probability allows people to look at every single option/perspective there is such as rolling a die. After going over the surveys in the lecture, I’ve come to understand that the given axiomatic theories must be understood completely. One small differential may result to a incorrect result which could lead to something such as overestimating the probability of breast cancer.
Your post was very inciteful and I liked how in-depth you went regarding the definitions. When you mentions differentials resulting in overestimations, I felt that was definitely something to think about. This is probably something that happens often in healthcare and it is truly sad. The prof mentioned that this happens because physicians fail to recognize the much larger group of people who are the exception ie; the pool of people who had cancer and tested positive. If conversations like these happened more frequent there may be more awareness to statistics that can aid in better diagnoses.
Krystal, notice that physicians ignored the fact that the pool of people who did not have cancer but tested positive is large (false negatives).
Hi Tahshin,
I completely agree with your emphasis on how axiomatic theories must be understood to the fullest to obtain accurate results. I also found it very interesting how every option can be weighed out by using the probability method, even in a healthcare setting. It is important that we spread awareness, understand conditional probability, and consider all groups of people for future diagnosis and treatment.
Watching the first video, I learned a lot about probability and the different rules, axioms. I feel I have learned probability in other classes but it had never been broken down this way so I appreciated the explanation One thing that caught my attention was the conditional probability. This is the probability that something occurs due to something else happening. The example that was used was the brain tumor being the cause of a headache and vice versa. I hadn’t known that probability was interchangeable based on what was most likely to happen.
The second video was a bit more confusing to me because of the equations. I had to rewind the video over again to understand how concepts could be written out to solve for numerical values. I also chose the wrong answer for survey 2 and probably would not be able to find the answer on my own if I had to.
Hi Krystal,
I also got the answer for survey 2 wrong. I found the videos helpful, and they help me understand my mistake. It is a strange topic for me, but I come to understand it.
I agree about your insight in probabilities, and I also agree how confusing it can be. I think explaining about probabilities is best done in physical class with step by step instructions on a blackboard.
After watching the two videos I felt like I understand more of the topic probabilities in the first video because it explains more about statistics, probabilities and the rules of axioms. As well Prof. Bayaz-Oturk explained the axiomatic rules related to probabilities. But watching the second video made me a little confused about probabilities and it’s rule.I wasn’t too sure of why she solved the question that way. However, Watching the two videos gave me a lot of information about probabilities and I was interested to know about it because we usually use them but I have never thought of that it got to be broken down to understand the concept of it.
I also got interested to know the conditional probability. I liked how you analysis of the conditional probability which you said “ This is the probability that something occurs due to something else happening.”which professor Bayaz explained that A happens lead to B happens .
I wonder what your response to the survey was. Did you overestimate the probability that the person has cancer given a positive test?
The professor gave an overview of probability and its accordance with axiomatic theories. I remember studying probability back when I took a statistics course. It was kind of a reminder/refresher in accordance with this topic. I personally don’t like this topic, but the professor’s overall explanation was good. At first, I was confused about the survey given. But after watching the professor’s videos, I have a better understanding. It’s important to understand the theories behind probability.
Hi Leila, I found statistic difficult in which you have to be very attentive.
I found Professor Gulgun Bayaz-Oturk’s presentation very helpful in understanding the relationship between probability theory and decision-making under risk. The one thing that really stuck out to me was the results from the gynecologist survey. It was very shocking to hear only 21% of physicians responded correctly, and 60% overestimated the probability of women diagnosed with cancer. I have learned how healthcare professionals do not apply conditional probability properly and fail to consider base rating factors. The overestimation was due to practitioners neglecting the pool of people who did not have cancer, but tested positive. They neglected the size of this large group in comparison to the group who were diagnosed with cancer and tested positive. In all, it is important to spread awareness and also educate ourselves on the use of probability in healthcare.
Hi Jordyn,
I totally agree with your takeaway from the gynecologist survey. When physicians are working so closely with such a scary thing, ie. cancer it is vital that they pay close attention to the use of probability and know exactly how many were diagnosed with cancer and tested positive. I never knew that probability played such an important role in healthcare but it was very interesting to learn and now have a better understanding for myself, to pay more attention to detail.
Thanks! Do you think that working with natural frequencies could be a solution to the base-rate neglect problem? This was suggested by Gerd Gigerenzer and his colleagues in an article. Even though I do think working with natural frequencies is easier, I do not see that students are more likely to choose the correct answer when provided natural frequencies instead of probabilities. I find this puzzling. I am not sure what was the case in your class and it would have been interesting to know.
The lectures provided by the professor talks about statistics in its probabilities which can get very confusing, but it is very crucial to understand when using probabilities in medicine because there are no certainties in medicine. With the use of probabilities, doctors can predict about a prognosis of a risky procedure or future outcome of a condition to its closest range. The predictions based on probabilities helps in decision making which is related to the case studies assigned in class. This lecture gives a very good presentation about probabilities associated with health conditions such as cancer which shows that statistics comes hand to hand with medicine.
Being able to apply statistics to medical procedure in hopes to reducing negative outcomes can come in handy. Higher risk procedure can lead to higher mortality rate with understanding probability of X number of patients surviving a certain procedure can help decide the risk factors involved.
Hello Benny,
I totally agree with your statements. It’s necessary to compute probabilities correctly to prevent misleading information. To add to that, this happens often through bias such as lead-time bias and overdiagnosis bias which seems to be a prominent issue in the U.S.
Kaitlyn, we also need to make sure that correct metrics are being used when comparing outcomes across different health care systems. Clearly, 5-year survival rates was not the correct metric to compare health outcomes in the U.S. and the U.K.
Hello Benny,
I agree with your statement. It can be very tricky trying to understand probability because it is very intricate, however, it does seem to a be a vital part of healthcare. It can be used to determine the likelihood of certain procedures and used in studies to develop life saving medicine. We can also use it to compare the success of various healthcare facilities to better predict what needs to be worked on.
The professor explains how statistics can be applied in the medical field to help us study medical condition outcomes. She goes into detail on how probability is used to help figure out the survival rate and annual mortality rate for prostate cancer patients. Now I understand how medical professions acquire these numbers which is quite interesting. It seems that research and statistics help determine the number of patients who are diagnosed and survive.
Hello Jason,
I was also intrigued on how statistics really do make a difference in the medical field. Surely enough, these statistics are the reason for things such as the covid vaccine to help the world like how it is today. You really do take surveys and data for granted until you realize the benefits of that actual analysis that goes in to it.
Professor Bayaz-Ozturk talks about an overview of probability and where it is used in finance, medicine, statistics, and engineering. The only experience I have with statistics is when I took a course for math. Statistics can be very confusing, it takes a lot of practice and you must pay close attention to detail. The first video helped me understand statistics since she broke it down so well. The second video stuck out to me when she gave the example about gynecologists. When she explained how 21% of them responded correctly, and 60% overestimated the probability of women diagnosed with cancer. It is shocking to me that they do not apply conditional probability correctly. This guest lecture helped myself become more educated with the use of probability in healthcare.
I find statistics confusing as well. The explanation of how to find the probability for scenario one made me recall my statistics course. At that time the instructor revealed the shocking truth of just how badly statistics can be misunderstood or poorly reported, just as Professor Bayaz-Ozturk did.
Statistical illiteracy can lead to overestimating or underestimating the likelihood of an outcome or event. This is an important thing to avoid in the medical field as patients will need to be informed of things such as success rate of procedures, probability of recovery or illness progression. An example discussed in the lecture was lead-time bias. Essentially, due to finding disease earlier than diagnosis the survival time of a patient is reported to be longer while the course of the disease remains unaffected. This type of bias can provide patients with a false sense of hope and unrealistic expectations. This can also make their passing harder on loved ones.
I agree that probability data opens up potential unfairness to certain topics when done poorly and having prior knowledge will help us become more effective as health care workers. The lecture questioned how I view the severity of diseases and that going forward we should be more cautious on how we report them.
The guest speaker discuss how probability descriptive theory which helps us to judge and make decisions. It was very import ant to know that using probability help to understand certain disease and there outcomes. The speaker makes it clear that some physicians may not be able to use statics when screening for cancer. The lecture enlighten the importance of statics in the health field and by not understand probability it can cause serious health consequence.
Professor Bayaz-Ozturk guides us on the overview of probability and decision-making under risk. In her first lecture, she goes into detail on probability with an example of six-sided dice and conditional probability. Alike, I remembered learning about all ranges of different probabilities including conditional in Statistics. It’s interesting to know that statistics are widely used in many fields of work. In the second lecture, she answered the surveys and explained them well through the mechanics of probability as I was confused with the survey choices beforehand. What was shocking to find out is through the experiment with 160 gynecologists, given the same survey, only 21% answered correctly. In addition, she stated that it was lower than the percentage of randomly guessing it correctly which is 25%, which may indicate some gynecologists were not even randomly guessing. This raises awareness of the importance of understanding and applying statistics in probability in the workplace.
Professor Bayaz-Ozturk gave us a lot of insight into probability and its relation to healthcare. While the first part of the lecture was about the techniques used in probability and their definitions the second part of the lecture was used to draw connections to healthcare. Personally, I found it very fascinating when healthcare statistics was explained using Gerd Gigerenzer theory. Lead time bias is probability based upon the number of patients still alive after a five year period divided by patients number of patients diagnosed with cancer. That would yield the survival rate, however, it just means that some people were diagnosed later while others were diagnosed at a younger age. Overdiagnosis bias is using both progressive and non-progressive cancer patients and seeing their mortality rate which would yield a much higher outcome due to both sets of cancer patients.
Actually, mortality rate would be a better metric in this context since it will not include non-progressive cancer cases (i.e., people will not die of cancer).
After watching the two videos I felt like I understand more of the topic probabilities in the first video because it explains more about statistics, probabilities and the rules of axioms. As well Prof. Bayaz-Oturk explained the axiomatic rules related to probabilities. But watching the second video made me a little confused about probabilities and it’s rule.I wasn’t too sure of why she solved the question that way. However, Watching the two videos gave me a lot of information about probabilities and I was interested to know about it because we usually use them but I have never thought of that it got to be broken down to understand the concept of it.
The solution made use of the Bayes’ rule presented in the first part of the lecture. Interpreting a medical test is an important application of the Bayes’ rule. Since we do not think of or apply this rule in everyday life, Gerd Gigerenzer was presenting an alternative presentation of data in the form of natural frequencies. Hopefully, that was easier to work with and so made more sense.
This series of videos have further emphasized the importance of probabilities in medical scenarios. The ability to take advantage of statistical data to perceive consistencies and help us form a medical resolution is an eye-opener for me. As such, what is interesting to me especially is the potential for biases in probability studies since data of this nature is meticulous. One of which was called the overdiagnosis bias that potentially results in inflation that influences the survival rate to be higher than it should be. I also find it intriguing that due to a country’s different approaches in treatment and potential biases, a more accurate report of their mortality rate instead of their survival rate is a much more preferred probability data to rely on.
It is interesting how the professor explains these different types of probability theories and how to apply them. I find it very interesting how the concepts of probability rules over how we make decisions. Whether we go red or black in roulette or play blackjack instead. After doing the survey and then watching the lecture it definitely made me understand better on how to make decisions based on specific probability theories.
Yes Richard I agree I found the video to be very interesting as well probability is very important and is a very useful tool when running a business and trying to make decisions in life.
The lecture gave an overview of cancer patients having the probability of cancer. They are using Statistics and probability to find the exact number of people who have cancer and their diagnosis. Also, it’s very interesting to how they are using probability in the medical field to fine the numerical from, what kind of cancer is affecting what percentage of people.
Probability is used in several disciplines to help make decisions in all aspects of Life probability can help Engineers health Finance and also can help form certain rules when it comes to trying to detour individuals in certain aspects of life. Probability can help predict certain outcomes which can help certain individuals or organizations or companies to see future results. When dealing with probability you have to try to think and figure out all possible outcomes versus actual outcomes which can help predict these of improvement or lacking of certain behaviors which can result in unwanted things happening. Probability includes all different types of formulas and equations. You can also look at different situations on how they relate to one another as well as the risk needed in order to accomplish. But statistics also include probability which is common among patients, doctors, politicians and journalists which is a language in itself.
I am happy to see that you found the lectures interesting and informative. The second lecture provided an example of a fallacy in probabilistic reasoning (i.e., base-rate neglect) and another example on statistical illiteracy. I wonder what your thoughts are regarding the way the information is presented? Do you think the approach suggested by Gigerenzer and his colleagues would lead to less errors in probabilistic reasoning? What was the result of the class experiment? Were the students more likely to choose the correct answer when information was presented in the form of natural frequencies rather than probabilities?
After you have read the article on cancer, you will see that use of survival rates in comparing health outcomes across different health care systems is a common mistake. I hope the second part of the lecture has been helpful in identifying the problematic areas in the article.