[Jessica Ramsey]

Looks like this episode will actually air...... tomorrow, and it will be Kari doing the "bigger boob" testing (picture from the trailer below)

[Angelus]

I'll just leave this here.

[Nimrod]

Of course I missed it -
How did it turn out and what was determined from their "tests"?

[Jessica Ramsey]

The episode as a whole was great, and they found out that Kari with small breasts and medium breasts (aka Kari's size) got the exact same tips, but when she became a DDD she was tipped 30% higher by men, and surprisingly 40% higher by women

[TheBoobGuy]

I wonder if they had tried a really big size (comparatively) like a GG, if she would have kept getting bigger tips, or if they would have started to drop because general public would have seen her as "a freak".

[Angelus]

Boobs.

[Palomine]

Wow, she's wearing a LOT of makeup. I guess this is a since-she-gave-birth development?

I TiVo'd the episode but haven't had a chance to watch it yet... I'm sure the whole ep is already around for DL, but frankly, I'm only interested in the Kari bits (pun intended ...shouldn't be more than a few minutes worth of excerpts at most). Has anyone seen a working source to view and save just the Kari/falsies stuff in HD? TIA!

[Nimrod]

I wonder if they had tried a really big size (comparatively) like a GG, if she would have kept getting bigger tips, or if they would have started to drop because general public would have seen her as "a freak".

I can not say if their method of testing really was constructed well enough to actually test that breast size has an effect on tipping. There are a wide range of variables still at play. Other aspects that could be working against accuracy include:

1) Tipping based on service
2) Tipping based on time of day
3) Tipping out of habit or reflexively (some people just do not want loose change in their pocket)
4) A one time random large tip - say $5 or $10

The first time I researched the early difficulty getting this episode going there were statements to the effect that one of the suggested tests was to have a series of servers (multiple women) change sizes over the course of several days - some getting bigger, others smaller, others jumping around in size and even one not changing at all.

This approach was supposedly dropped for two reasons:
- It was more involved in staff, time, cost and potential impact on a business willing to act as the test location.
- It was almost exactly a test outlined in the concepts of the plaintiff raising legal challenges to this episode.

However - this would have been a far superior test, yielding many more data points, not influenced by a new server's skills improving (like Keri) and it even had a control built in. From my digging the producers of Mythbusters did start to set up a test along these lines but scrapped it early on which could have led to the comments about scenes being cut out and the focus changing around Kari.

Abstract

Taken from: Determinants and Consequences of Female Attractiveness and Sexiness: Realistic Tests with Restaurant Waitresses by Michael Lynn -
Archives of Sexual Behavior
October 2009, Volume 38, Issue 5, pp 737-745

Waitresses completed an on-line survey about their physical characteristics, self-perceived attractiveness and sexiness, and average tips. The waitresses’ self-rated physical attractiveness increased with their breast sizes and decreased with their ages, waist-to-hip ratios, and body sizes. Similar effects were observed on self-rated sexiness, with the exception of age, which varied with self-rated sexiness in a negative, quadratic relationship rather than a linear relationship. Moreover, the waitresses’ tips varied with age in a negative, quadratic relationship, increased with breast size, increased with having blond hair, and decreased with body size. These findings, which are discussed from an evolutionary perspective, make several contributions to the literature on female physical attractiveness. First, they replicate some previous findings regarding the determinants of female physical attractiveness using a larger, more diverse, and more ecologically valid set of stimuli than has been studied before. Second, they provide needed evidence that some of those determinants of female beauty affect interpersonal behaviors as well as attractiveness ratings. Finally, they indicate that some determinants of female physical attractiveness do not have the same effects on overt interpersonal behavior (such as tipping) that they have on attractiveness ratings. This latter contribution highlights the need for more ecologically valid tests of evolutionary theories about the determinants and consequences of female beauty.

This does not follow the results of the Mythbusters Bigger is Better testing. They did not have a linear result - both the small and medium sizes got similar tips and only the larger size had a positive jump. Given the history of this type of testing the medium size should have been much closer to a mid-point between the small and big results. Thus it is the Mythbusters test that is called into question as it does match up with the body of research - they are the exception. Their test could have been much better constructed.

And to throw another wrench into the mix we must also consider the Kari may have inadvertently affected the test because she could have subtly changed her actions when she was "modified" in both the small and large states. Her patrons may not have known she was changed but she did. It is totally possible that her small and large sizes were tipped atypically more and that her natural medium size actually is her normal natural tip rate.

Addendum:
Attached small GIF movie of the BE animation used in this episode (Episode 220)

[Angelus]

Wow, she's wearing a LOT of makeup. I guess this is a since-she-gave-birth development?

No.  Without watching the episode, I'm guessing it's either 1) a disguise component (see: Wig) or 2) Get people to look at her boobs rather than her face.

[Jessica Ramsey]

Wow, she's wearing a LOT of makeup. I guess this is a since-she-gave-birth development?

I TiVo'd the episode but haven't had a chance to watch it yet... I'm sure the whole ep is already around for DL, but frankly, I'm only interested in the Kari bits (pun intended ...shouldn't be more than a few minutes worth of excerpts at most). Has anyone seen a working source to view and save just the Kari/falsies stuff in HD? TIA!

She is wearing the extra make-up to help prevent people from recognizing her.

Pal replies: ah, that makes sense I suppose. I thought maybe she'd been seeing Krusty the Clown on the side or something.

[Theodore]

I can not say if their method of testing really was constructed well enough to actually test that breast size has an effect on tipping. There are a wide range of variables still at play. Other aspects that could be working against accuracy include:

1) Tipping based on service
2) Tipping based on time of day
3) Tipping out of habit or reflexively (some people just do not want loose change in their pocket)
4) A one time random large tip - say $5 or $10

The first time I researched the early difficulty getting this episode going there were statements to the effect that one of the suggested tests was to have a series of servers (multiple women) change sizes over the course of several days - some getting bigger, others smaller, others jumping around in size and even one not changing at all.

This approach was supposedly dropped for two reasons:
- It was more involved in staff, time, cost and potential impact on a business willing to act as the test location.
- It was almost exactly a test outlined in the concepts of the plaintiff raising legal challenges to this episode.

However - this would have been a far superior test, yielding many more data points, not influenced by a new server's skills improving (like Kari) and it even had a control built in. From my digging the producers of Mythbusters did start to set up a test along these lines but scrapped it early on which could have led to the comments about scenes being cut out and the focus changing around Kari.

This does not follow the results of the Mythbusters Bigger is Better testing. They did not have a linear result - both the small and medium sizes got similar tips and only the larger size had a positive jump. Given the history of this type of testing the medium size should have been much closer to a mid-point between the small and big results. Thus it is the Mythbusters test that is called into question as it does match up with the body of research - they are the exception. Their test could have been much better constructed.

And to throw another wrench into the mix we must also consider the Kari may have inadvertently affected the test because she could have subtly changed her actions when she was "modified" in both the small and large states. Her patrons may not have known she was changed but she did. It is totally possible that her small and large sizes were tipped atypically more and that her natural medium size actually is her normal natural tip rate.

She worked the same shift, I do not know if it was the same day of the week (meaning three weeks, or the same shift three days in a row) ... the shift however was the same they even stated that the time of day was controlled.  Kari literally stated "I will work the same shift."    A random large tip is irrelevant to the test and does not need to be control for because one there is no way to control for it and a person who gives a large tip will either always give it or give it to the person he is most attracted too (well within the acceptable testing for this.)   However if they wanted to keep track, they would need to keep track of individual tips rather than the whole amount.   Also, there is no reason to expect a linear relationship to prove the myth that is a silly expectation (for example, it could be exponential relationship or be flat at the lower sizes).   All that matter was a noticeable jump from the smaller sizes to the larger and they showed that. 

Your other complaints are well valid ... and you ignored that even though she was made up, people may have recognized her and didn't say anything knowing they may be testing something at that time.   And your observation on Kari is spot on, she was herself when normal, was "in pain" when small and appeared to be more flirty when large.   

[Nimrod]

She worked the same shift, I do not know if it was the same day of the week (meaning three weeks, or the same shift three days in a row) ... the shift however was the same they even stated that the time of day was controlled.  Kari literally stated "I will work the same shift."

Yes - she worked the same shift so the hours of the day she worked were the same - however is appears the days were consecutive so the day of the week was different and that would have thrown their data off even more.
 
A random large tip is irrelevant (why is that?) to the test and does not need to be control for (Again why?) because one there is no way to control for it (There most certainly is) and a person who gives a large tip will either always give it (total statement of assumption) or give it to the person he is most attracted too (what about for good service, extra change, pay day generosity, sympathy, or other reasons?), well within the acceptable testing for this.

A single $5 tip would have been a 7% spike in the data and given that they were dealing with such small amounts and a single session of collection that is a significant potential for the introduction of a misleading spike. Remember they found the men only changed on the 'big' day by +30% and women +40%.

However if they wanted to keep track, they would need to keep track of individual tips rather than the whole amount.

They did keep track of who gave, their gender and the amount - they even commented on the specific amounts at times so they knew.

Also, there is no reason to expect a linear relationship (Actually that is why I included examples of previous research about this very topic showing that there is a propensity for a linear relationship and not exponential, logarithmic, parabolic, random, etc.) One would expect a linear relationship given the body of work - their results were not typical) to prove the myth that is a silly expectation - for example, it could be exponential relationship or be flat at the lower sizes. (No, that is not what the previous research tests have shown)  All that matters was a noticeable jump from the smaller sizes to the larger and they showed that. (Ah, they did not show that)

What they showed is that when you move from small to medium size you get no change at all in tip amount - they got the exact same amount. This the not at all what would be expected. When they moved from medium to large they did get an increase which is as expected however that was essentially only half the test results and there are a number of reasons already given why their testing is not very conclusive.

At the very start of the segment they repeatedly stated that they would test if by just changing breast size they would get an increase in tips. When Keri increased by two cup sizes she got exactly the same tips. This is not showing that when you increased to a larger size your get a noticeable jump.

Your other complaints are well valid ... and you ignored that even though she was made up, people may have recognized her and didn't say anything knowing they may be testing something at that time.   And your observation on Kari is spot on, she was herself when normal, was "in pain" when small and appeared to be more flirty when large.

The reason I did not mention the whole "people might recognize her" point is because I had already made that exact point back near the beginning of this thread topic. Just the fact that they used Kari was a poor choice. Also, if their tests were ran on consecutive days as it appears, then repeat patrons may have noticed something odd about their server. And, it is totally possible that by the third day that word had leaked out that Mythbusters was in the area and the tip spike on the third day had more to do with **29** than cleavage.

I should point out that if one looks at the whole process they used to test this there are many basic problems:
1) Very small sample size
2) A single test subject
3) No true control point
4) Multiple data contamination potentials
5) Ignoring contrary results to established norms
6) Possible test subject bias or manipulation

Also the complete dismissal that women actually gave more tips for the large size was given a few seconds and not even commented on. Why was that? That alone brings up more questions and was just tossed in at the end... very disappointing.

This episode segment was given the title "Storm in a D cup" which is a variation of "Tempest in a Tea Pot" given that it was at a coffee shop and probably because they felt that some viewers would not understand that a tempest is a storm and that the saying deals with making a big deal out of very little- however they are still in error because the only size that actually made an impact was an F cup...
So they were still off the mark.

[Theodore]

First, do not get on my case for assuming and then make assumptions yourself.   There is no way to know whether or not the were same day of the week or each day one after the other, it doesn't appear to be consecutive days to me at all.  In fact, given the show it is most likely not consecutive days what so ever as they aren't just working on one thing at a time so there is likely a gap between the days.

I love how you go "Why is that?" "again why?" without getting to where I explained why.  As for the reason you list those fall under the "always give it" the people who tip well for good service always tend to tip well for good service.  People who tip well always do for the same reasons ... seriously, you want to list them separate reasons but in the sense of the test they aren't.   You call it assumption, but it isn't ... in fact the only reason why I listed the most attracted too as separate is because it is something being tested for.   You are calling for a control because it can be a large swing  ... irrelevant.   They are testing what gives more tips, so size of an individual tip is not relevant.   The issue here is they should have done multiple days at the sizes and gotten averages for the sizes as opposed to just one day for each size, this still wouldn't not need to control for tip size.  And btw, they didn't keep track of individual tips amounts ... in fact, on normal and small they never stated an individual tip amount.   In fact, the only amount they stated was "two dollars" on the large day which wasn't even right as he put in change as well (as heard by when he dropped in the first dollar).   There is zero evidence to support them keeping track of individual tips.

I am sorry, they didn't show Kari getting larger tips at a larger size?  Gee, I must not have watched the same thing you did.   Again, you can expect two sizes to be similar, and a lot of "linear" results aren't linear at all.   Sometimes they do a line of "best fit" ... and guess what?  You can do a line of best fit with results like they got   They would need to explain why the smallest size is higher than expected and normal was lower (which could be that it was Kari's first day), but that is it.  Their results do not deter from a linear result in other testing at all.  

Multiple subject doesn't improve results as much as you think, it helps but adds it far more variable that need to be controlled.   A single subject can be more easily controlled, also we are dealing with a coffee house here ... at most you are talking about is maybe around three servers (could be two or up to five, but I have personally not encountered shops with more than five).   So if you use all the servers in the test you are more likely to have people figure out something is going on rather than just one person in the test.  What they should have done is one server for three days each week and either did one size for the entire week (or change sizes ie Mon Normal Wed Small and Fri Large and then switch on later weeks.)    In order to have more people, they would need something like a restaurant to use and that throws in even more variables that need to be controlled.

[Nimrod]

Theo does raise some good counter points for the need of a control based on the reasoning that what was being tested was whether or not a larger bust line would generate larger tips. However my assertions are that there were too many variables and results that were either not addressed of simply ignored that should have allowed them to make the confirmed conclusion at the end of this test. Simply put, the one shift Kari was augmented more tips were put in the jar - if this had to do with her bust size is possible but not definitive. Here is why:

Proximity to M5 workshop
The Cafe Roma coffee house is in the same neighborhood as the M5 workshop where Mythbusters is produced. It takes less than 4 min. to drive door-to-door and they are about 9000ft apart. It is conceivable that people in the area are familiar with Kari enough to at least think they recognize her given the amount of on-site, exterior and local shooting they do and the fact that she has been in the area for at least 15 years. The very fact that they were wanting a large sample size by going to a local coffee shop actually increases the chance that a patron would recognize her.

Using Kari as the single test subject
I have explained why this was a poor choice. Not only for the recognition factor, all the higher due to the proximity to M6 but because she is not a barista and therefor her service would have been very inconsistent. Giving her the benefit of doubt we can assume she improved as the days went on. But this would have positively affected her tips simply because she got better and increases the chance that her tips would be service based. Not to mention that they even say "Kari is really working it." on a modified day which shows that she was changing her interactions - again thrown off the test.

Determining days of testing - the impact of Friday
This is very simple to know - just look it up. This Cafe Roma (it is a small chain) is closed on the weekend and the crew set up the cameras, monitors and such on the first weekend in December, 2013 and filmed some small segments when it was closed and vacant. They did not film segments or test Monday in order to see if patrons were suspicious, test their set-up and because this is one of the slower times and sample size would have been low over lunch. The actual test days were the last three days ending on Friday. Because Friday is the second best tip day (behind Saturday) it is probable that the spike in tips on the last day, the "big" day was because it was Friday. This could also help explain why the "small" day equaled the "normal" day as it was filmed on the Thursday. Monday is the worst tip day of the week days, Friday the best.

Not having a true control
They needed to have at least a 2nd "normal" size barista working concurrently also getting tips. This would have let them at least compare two workers at differing sizes both at the same time and would have lessened the impact of quality of service, flirtation, a high tip day, etc. They would have had another value to make a relativistic determination at least. A single data point at a single size for a single person is just a stab in the numeric dark.

Keeping track of the tips
To say they did not know the amount of tips and then to acknowledge that they knew the amount of the first tip ($1) and the $2 tip during the "big" phase shows that they could and did know some of the tips. They even had a tip jar camera. They are shown in the show tracking patron number, patron gender, how they interacted with Kari and in at least 2 examples shown they knew the amount of the tip. So, one can not say they did not know any of the tip amounts, they would have been aware of at least disproportionate ones.

Ignoring the larger "big" day tip amount by women
This is completely counter-intuitive and the fact that it occurred on the last day may actually reinforce the fact that later in the week tips increase naturally. They literally spent 7 seconds talking about this odd result and then just drop it. They also did not give the amounts given by women on the "normal" or "small" days. This would have at least helped - but as they were not given one could assume that they mimicked the male results.

Other research tests on this very topic have shown that there is a linear relationship between bust size and amount of tip. The Mythbusters did not get a linear relationship so their results are more suspect. Regardless of the failing set up of their test and taken just within itself this test does not follow the expected results - the "small" size exactly equaled the "normal" and women patrons gave much more than the men on the "big" day.

[Angelus]

Can we just agree that this was just a ratings stunt to show Kari in a stuffed bra?  

I certainly agree with that statement. -Pal