­The table shows the sales forecast from January to June of last year

­Further explanation of this upon request:

A "linear relationship" is a line. When I say x and y have a linear relation, it represents a line on the xy axis. i.e. y = mx + c

I do not understand what is meant by a "constant linear relationship". I would just take it as a line i.e. a linear relationship.
I would not think that it means y = mx. As I mentioned before, if there is some particular curriculum in which a constant linear relationship means "y = mx" then I do not know. But a google search did not give any such information. So I doubt many people would equate "constant linear relationship" to y = Constant * x and hence the question would not pass the GMAT filter.

We know that y = constant * x is direct variation and I would look for that term to arrive at the conclusion that I have to use that relation here.

Hence, in this question, as given, this is what I imagine is also possible:

Hence it is possible that the actual number of units sold in April is less than the number sold in March even using the data from both statements.

Hence my answer here would be (E).
­

­Further explanation of this upon request:

A "linear relationship" is a line. When I say x and y have a linear relation, it represents a line on the xy axis. i.e. y = mx + c

I do not understand what is meant by a "constant linear relationship". I would just take it as a line i.e. a linear relationship.
I would not think that it means y = mx. As I mentioned before, if there is some particular curriculum in which a constant linear relationship means "y = mx" then I do not know. But a google search did not give any such information. So I doubt many people would equate "constant linear relationship" to y = Constant * x and hence the question would not pass the GMAT filter.

We know that y = constant * x is direct variation and I would look for that term to arrive at the conclusion that I have to use that relation here.

Hence, in this question, as given, this is what I imagine is also possible:

Hence it is possible that the actual number of units sold in April is less than the number sold in March even using the data from both statements.

Hence my answer here would be (E).
­

Given the brevity of information provided the answer to this question should be (E). I don't know why the questions has not been corrected when so many users have pointed out an obvious flaw.

How can it be E?
You need k to be -ve to have a slanting slope right? Y=kx+c

Even if k is between 0 and 1. The values are reducing in same direction.

But, these are items and actual and predicted can’t have a negative relationship.

Maybe I’m missing something obvious here. Bunuel could you please weigh in?

They for sure can if the negative relationship is bound to the first quadrant in a graph. This is possible.

Adarsh_24
Mate it could be the following case: January predicted is 12 and sold is 15, February predicted is 16 and sold is 14, March predicted is 16 and sold is 12, April predicted is 20 and sold is 9 ,May predicted is 20 and sold is 6, June predicted is 24 and sold is 2.
In this given case none of the values are actually negative however if you were to use these values to compute the gradient then it would indeed be negative. Also notice how all these values would be negatively correlated but also in the first quadrant. This is completely possible as statement 2 does not say if the relationship is positive and linear. Hence, you can't just guess that it is.

­No,

Number of Products sold can never be negative.

­No, this wording and context implies a positive correlation, where Actual = (Some Constant) * (Predicted)

­
The table shows the sales forecast from January to June of last year, by the sales team of Company X, for the number of units of Product Z each month.

Was the actual number of units of Product Z sold in March less than the number sold in April?

(1) Twelve units of Product Z were sold in March.

(2) From January through June of last year, there was a constant, linear relationship between the forecast and actual number of units of Product Z sold.­

­

­No, this wording and context implies a positive correlation, where Actual = (Some Constant) * (Predicted)