MGMAT 1 CR

[ShobhitK282]

Joanne: An increasing number of online retailers now allow customers to create "wish lists" of items they would like to receive as gifts. Such lists are certainly useful, but these retailers should also explore other ways to suggest gift purchases for these customers. In particular, without revealing the specifics of a customer's purchase history, a website could quickly analyze a retailer's entire inventory, select a list of items similar to those the customer has already purchased, and then e-mail that list to a group of contacts specified by the customer. Such a system would suggest gifts that, because of their similarity to the customer's prior purchases, would be extremely likely to appeal to the customer. In the argument, Joanne assumes that the hypothetical customers

(A) are familiar with most or all of the items in stock on the websites where they shop
(B) would prefer novel gifts that are unlike the items they currently own
(C) do not use retail websites primarily to purchase gifts
(D) would be relatively unconcerned if their retail purchase histories were available to others
(E) prefer online shopping to shopping in physical retail stores

I'm confused between OA(C) and E? I chose E because if we negate it-doesn't prefer online shopping. This would mean that her history is not indicative and hence the gift would nout appeal to her. Why is my reasoning wrong?

[jnelson0612]

Hi Shobkit,
Let's break it down:
Conclusion: Gifts bought through this system would be likely to appeal to the recipient.
WHY?
Premise: The online retailer would analyze a list of the recipient's past purchases, and offer a list of similar items to the person wishing to purchase the gift.

What is being assumed?
That the recipient's purchases are actually things for the recipient, that the recipient would enjoy receiving as a gift. For instance, if I use my online account to buy things for my business, such as notebooks, staplers, scotch tape, etc., then this potential gift list would contain similar office supplies. Do I really want those as a gift from someone? No. Or maybe I've just been shopping online for my daughter, purchasing Barney videos and Playdoh. I really don't want to get similar items as a gift for myself. :-)

So you must assume that the things a person buys online actually reflect things that that person would like to receive as gifts.

C) is our answer. If these people are using websites to purchase gifts for others, than similar items would not necessarily appeal to these people as gifts for themselves.

E) is pretty irrelevant. Whether people prefer to buy online or in-person doesn't matter a lot because you could presumably get similar items using either method. What does count is whether, when people do buy online, they are buying things they would actually like to receive as gifts.

[Steven Zheng]

Actually it is a statistics problem.
To make it more interesting, Let's assume the retailer in the argument is Amazon.
Amazon wants to know what kinds of goods a customer is actually prefer. So he picks out a sample of the shop list to "guess" what the customer is really like to buy.
However, here come to 2 questions.
First, is the sample large enough to indicate the customer's preference? For example, if I bought 9 apples and 1 knife on the Amazon and he only picked out "1 sample" and fined out "Oh, this man likes knife because he has bought one!" That would be a mistake because actually I bought more apples than knifes which "can" mean I like apples more than knifes.
Second, does the sample really indicate the customer's preference? That means, the 9 apples and 1 knife in the list are the things I really prefer to buy, rather than other reasons such as I bought it for someone else or I bought it because there was an emergency.

Now let's come back to the beginning-Why I have to clarify it is a "statistics" problem?
Because when you find out it is a "sample to population" problem, you only need to concentrate on the sample to check whether the indication of the sample is flawed, which can only happen in 2 ways, the sample is not large enough to represent the whole thing or the sample is not abundant enough to represent the whole situation.

And on the basis of statistics, there is only one choice related to the "sample" question, choice C, which question the abundance of the sample.

PS:the abundance here means the sample can represent all possible preference, not means the sample includes a variety of goods. That's different.

Hope my poor English helps :)
I'm a Chinese.
Awkward English, though I've tried my best.

[RonPurewal]

Actually it is a statistics problem.
To make it more interesting, Let's assume the retailer in the argument is Amazon.
Amazon wants to know what kinds of goods a customer is actually prefer. So he picks out a sample of the shop list to "guess" what the customer is really like to buy.
However, here come to 2 questions.
First, is the sample large enough to indicate the customer's preference? For example, if I bought 9 apples and 1 knife on the Amazon and he only picked out "1 sample" and fined out "Oh, this man likes knife because he has bought one!" That would be a mistake because actually I bought more apples than knifes which "can" mean I like apples more than knifes.
Second, does the sample really indicate the customer's preference? That means, the 9 apples and 1 knife in the list are the things I really prefer to buy, rather than other reasons such as I bought it for someone else or I bought it because there was an emergency.

Now let's come back to the beginning-Why I have to clarify it is a "statistics" problem?
Because when you find out it is a "sample to population" problem, you only need to concentrate on the sample to check whether the indication of the sample is flawed, which can only happen in 2 ways, the sample is not large enough to represent the whole thing or the sample is not abundant enough to represent the whole situation.

And on the basis of statistics, there is only one choice related to the "sample" question, choice C, which question the abundance of the sample.

PS:the abundance here means the sample can represent all possible preference, not means the sample includes a variety of goods. That's different.

Hope my poor English helps :)
I'm a Chinese.
Awkward English, though I've tried my best.

this is not the point. in fact, nothing in the passage suggests that the proposed system considers only a statistical sample; the actual wording-- "items similar to those the customer has already purchased"-- actually suggests the opposite, that the algorithm is considering the customer's entire purchase history.

the point is, as jamie (jnelson0612) explained above, that there's no way of knowing whether the customer's purchases have been for him/herself, rather than for someone else (in which case they wouldn't reflect the customer's own preferences at all).

[ElizabethS105]

What is confusing me is this line: "... and then e-mail that list to a group of contacts specified by the customer." This makes me think that the list is actually for other people and not the individual. Am I thinking about that line wrong?

[krutiks511]

the same line confused me too...please explain this line,that would be a great help.

[Alok chandraS808]

Hi Can someone explain that line in the comment mentioned above mine? That line states that the list is sent to friends and friends will select what has to be decided to be sent as a gift to the customer. If it is the case, If i purchased inner ware from CK,or a toilet cleaner, I do not want it to be revealed to my friends and I do not want them to see what all I purchased. In that case opt D is correct.

[Sage Pearce-Higgins]

"a website could quickly analyze a retailer's entire inventory, select a list of items similar to those the customer has already purchased, and then e-mail that list to a group of contacts specified by the customer"
The idea is that the retailer could analyse a customer's purchase history and find similar items that the customer might like to receive as gifts. Such a "gift-list" could be sent to the customer's friends. However, as jnelson pointed out, the system would fail if the customer wasn't buying items for herself, but for gifts for other people: i.e. the customer might not want to receive similar items to the ones that she bought. It's actually a real-world problem that retailers have in understanding their customers' preferences and making marketing decisions based on their data.

[JbhB682]

Hi - not sure i understand why D is irrelevant per the solution stated

Negating D, customers are concerned about releasing information to others (Others in my thinking are obviously, friends of the customer who will buy the gift for the customer but also others can mean : corporations)

I understand that the online retailer will not release the exact purchase to the friend of the customer BUT the online retailer will review past inventory of the customer in order to make the suggestion ( the corporation will certainly have access to what the customer has bought)

So the question is : how can we be sure that customers are not concerned about releasing information to others (specifically corporations)

Thank you !

[Sage Pearce-Higgins]

Good thinking that "others" could be talking about corporations as well as individuals. However, it sounds like you're getting a bit too picky here. Presumably the website could analyse the data without actually showing it to any person, so it could pass a confidentiality test.

However, this is all a bit of a distraction. In any case, the conclusion could still work: "Such a system would suggest gifts that would be extremely likely to appeal to the customer." Be sure that you're addressing the actual conclusion and not rating the plan in a general way.

[sahilmalhotra01]

Hi MGMAT Expert,

I have doubt regarding the option choice A.

Option Choice A : are familiar with most or all of the items in stock on the websites where they shop

Negated version of choice A : are not familiar with most or all of the items in stock on the websites where they shop

If the customers are not familiar with most the items in stock on the websites, then how the gift recommendations from these items appeal to the customers.

So, I understand that option choice A is an assumption.

Please help me in understanding how the option choice A is incorrect.

Thanks

Sahil Malhotra

[Sage Pearce-Higgins]

Good application of the negation test, but I think you're making an assumption. Your logic is that we need to be familiar with a product in order for it to appeal as a gift. I don't think this is true! Surely we can like things that we're not familiar with, and maybe the best presents are surprises.