Ambrose Li Jun 15, 2011:

To paraphrase the paragraph,

50% of all travellers – stay less than 4.0 days
50% of all travellers – stay more than 4.0 days
Holiday travellers – stay on average 6 days
“Education segment” – stay on average 125 days
All travellers – stay on average 18.5 days

You really don’t know how many days each type of traveller actually stays in NZ. You only know an average.

If we try plotting these graphs, we might expect the graph of each type of traveller to be roughly symmetrical. However, if we try plotting the graph for all travellers, the “hump” of the graph would be on the far left with a long tail to the right (and possibly a secondary hump on the right somewhere): It will be a very asymmetrical distribution. This is the “skew” that the text is talking about: The long stay of students is “skewing” the average to 18.5; it is a skew in a strict mathematical sense.

PS: I hope my description of the graphs in the last paragraph makes sense. Statistics has never been my strong suit and I could have mixed things up.

Jinhang Wang Jun 15, 2011:

理解的关键是 skew 一词，即偏斜。这里每类旅行者的比例差异较大，即所占的权重不同。

Ambrose Li Jun 15, 2011:

I believe “median” and “average” (i.e. “mean”) are used here in their strict mathematical sense.

“median length of stay in New Zealand was 4 days” = 50% of Chinese visitors stayed less than 4 days, 50% of Chinese visitors stayed more than 4 days

“average length of stay is 18.5 days” = The total number of days Chinese visitors stayed in NZ divided by the total number of Chinese visitors is 18.5

albertdeng (asker) Jun 15, 2011:

Does "education segment" simply mean international students? Who stays 18.5 days - visitors from all countries or all visitors from China?