Jump to content
Sign in to follow this  
Vishwadeep Khatri

Bench and Mark 3 - Men vs Women - Who Won?

Recommended Posts

Check this debate between Bench and Mark on announcing winners for a contest. Is Bench incorrect here? 

 

Men vs Women Who Won..PNG

 

Description - There is an announcement that the results for Men vs Women contests are about to be announced by Bench an Mark. The reward is to be given to better performing gender. To add spice to the contest, men and women had been divided into younger and older folks. 
 

Bench comes on stage and says that he will announce the result with some analysis. He explains that younger men have done better than younger women. He then goes on to show that older men have also done better than their older counterparts. With this analysis, he concludes that men have won the contest. Dramatically, Mark comes in and announces that the final winners are women. Bench argues with Mark saying that men were better in both the categories. Mark says that overall percentage pass rate of women is better than that of men and shows the aggregate scores. Mark highlights that the award was meant for better overall performance. Bench cannot figure out how this makes any sense. Bench continues to think that if there are only two categories possible and men did better in both, men should be considered winners. 

 

The cartoon highlights a possibility of misrepresentation that exists in many analyses, especially where sub-grouping is done and subgroups are analysed. 

 

Feel free to discuss this below. 

Share this post


Link to post
Share on other sites

Hahaha!! Bench is a victim of not knowing Simpson's Paradox :) 

 

OR 

 

Both Bench and Mark are victims of lack of Operational Definition for winning criteria :)

 

PS - I got to know about Simpson's Paradox from the 2 weekly questions that are asked. So thanks for it. For the uninitiated, go to Forum Dictionary and search for Simpson's Paradox!!

 

 

 

Share this post


Link to post
Share on other sites

absolutely true, i do agree with Natwar Lal that operational definitions were not clearly and meticulously defined, thats why the actual results varied from what was observed. 

Here the criteria for rational subgrouping does not clearly reflect the intent of subgrouping. 

moreover subgroups should be placed in the order in which they were collected otherwise this will definitely reflect our analysis. 

data collection plan should be clear and precise to our study.

Share this post


Link to post
Share on other sites
Guest Rakesh Sharma

I agree that Operational Definition was not defined.However while performing analysis it's first important to study and understand data..Here percentage is applied for analysis which is a derived data...Hence we need to go to source data and validate it...I feel a better analysis would have been by considering median..

Share this post


Link to post
Share on other sites
38 minutes ago, Guest Rakesh Sharma said:

I agree that Operational Definition was not defined.However while performing analysis it's first important to study and understand data..Here percentage is applied for analysis which is a derived data...Hence we need to go to source data and validate it...I feel a better analysis would have been by considering median..

 

Hi Rakesh, you are right in saying that actual scores should have been used. As this analysis is utilizing a pass-fail data, the option was to look at pass percentage. If actual scores were available, it would have been possible to consider median. 

Share this post


Link to post
Share on other sites
2 hours ago, Natwar Lal said:

Hahaha!! Bench is a victim of not knowing Simpson's Paradox :) 

 

OR 

 

Both Bench and Mark are victims of lack of Operational Definition for winning criteria :)

 

PS - I got to know about Simpson's Paradox from the 2 weekly questions that are asked. So thanks for it. For the uninitiated, go to Forum Dictionary and search for Simpson's Paradox!!

 

 

 

 

Absolutely Natwar, Simpson's Paradox it is. 

Share this post


Link to post
Share on other sites
53 minutes ago, shashankparihar19 said:

absolutely true, i do agree with Natwar Lal that operational definitions were not clearly and meticulously defined, thats why the actual results varied from what was observed. 

Here the criteria for rational subgrouping does not clearly reflect the intent of subgrouping. 

moreover subgroups should be placed in the order in which they were collected otherwise this will definitely reflect our analysis. 

data collection plan should be clear and precise to our study.

Hi Sashank, 

 

Subgrouping is usually not done with discrete data unless there is a sound objective behind it. If overall scores was the winning criteria, looking at groups must have been avoided. You have mentioned correctly that the intent of sub-grouping is unclear.  

Share this post


Link to post
Share on other sites

I believe sampling (number of attempts) should be common for both the teams, since at the end of the day if the overall numbers are clubbed, the probability of winning  changes due to the number of attempts.

This reduces variations in study and result will be acceptable.

 

Attempts and goals should be common in competition.

 

Share this post


Link to post
Share on other sites
On 10/30/2019 at 12:52 PM, Vishwadeep Khatri said:

 

Hi Rakesh, you are right in saying that actual scores should have been used. As this analysis is utilizing a pass-fail data, the option was to look at pass percentage. If actual scores were available, it would have been possible to consider median. 

median analysis is best suited when we have outliers in our collected data.

 

Share this post


Link to post
Share on other sites
On 10/31/2019 at 1:07 AM, Rahul PS said:

I believe sampling (number of attempts) should be common for both the teams, since at the end of the day if the overall numbers are clubbed, the probability of winning  changes due to the number of attempts.

This reduces variations in study and result will be acceptable.

 

Attempts and goals should be common in competition.

 

Hi Rahul, let us assume this to be population data. As number of men and women taking part in a contest are not same, imagine that the judges took a decision that winning gender will be one that has bigger percentage success rate. 

Share this post


Link to post
Share on other sites
Guest Kingsley Gibbs
On 10/28/2019 at 7:42 PM, Vishwadeep Khatri said:

Check this debate between Bench and Mark on announcing winners for a contest. Is Bench incorrect here? 

 

Men vs Women Who Won..PNG

 

Description - There is an announcement that the results for Men vs Women contests are about to be announced by Bench an Mark. The reward is to be given to better performing gender. To add spice to the contest, men and women had been divided into younger and older folks. 
 

Bench comes on stage and says that he will announce the result with some analysis. He explains that younger men have done better than younger women. He then goes on to show that older men have also done better than their older counterparts. With this analysis, he concludes that men have won the contest. Dramatically, Mark comes in and announces that the final winners are women. Bench argues with Mark saying that men were better in both the categories. Mark says that overall percentage pass rate of women is better than that of men and shows the aggregate scores. Mark highlights that the award was meant for better overall performance. Bench cannot figure out how this makes any sense. Bench continues to think that if there are only two categories possible and men did better in both, men should be considered winners. 

 

The cartoon highlights a possibility of misrepresentation that exists in many analyses, especially where sub-grouping is done and subgroups are analysed. 

 

Feel free to discuss this below. 

 

Share this post


Link to post
Share on other sites
Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

×
×
  • Create New...