The conclusion of this argument is - "if servers in Canada regularly wrote “Thank you” on restaurant their average income from tips would be significantly higher than it otherwise would have been". And the reasoning to reach this conclusion is stated as- "tips on these bills (the bills randomly generated with "thank you" message) were an average of three percentage points higher than tips on bills without the message". Now we need to look for an option that
must always be true for the conclusion to be valid.
A. The original argument does not talk about the category of customers who pay these tips, so this option does not seem relevant
B. This option definitely affects the validity of the conclusion. If the patrons find the "Thank you" message a regular practice then they may revert to older tipping habits. This would decrease the average income from tips (conclusion). Hence this is
correct choice
C. This information might very well be true but does it impact the conclusion in any way? Let's say it reminds customers that tip is a very significant part of income of servers but this does not mean that the customers are obliged to pay tip. Hence this is not relevant.
D. This option brings a condition that rate of tipping does not vary with how expensive the restaurant is. But it does not carry enough information to affect the conclusion in any way. Hence the option is incorrect.
E. This option talks about those patrons who, as a part of the study, were randomly given a "Thank you" message in the bill. This is very similar to the reasoning (premise) given in the argument to reach the conclusion and does not bring in any new information. Hence, this choice is incorrect.