<On some estimates by a green group>
Ok, let me keep it brief.
1. In estimating a figure from a set of samples to represent the whole population, you need to be very careful in making sure that your set of samples are representative and reliable. There are many ways to do sampling (wiki here), but generally, the more samples you have, the more reliable and convincing estimation you have.
2. Today some headlines like the following catches my attention :
The content is similar to the following (the following is from mingpao):
綠色力量在本月初訪問300多名市民 (should be households)，受訪家庭去年平均丟棄0.81個月餅，較對上一年減少3成半。另外，5成半的受訪家庭，會將月餅盒送往回收，數字上升1成7。
雖然受訪者丟棄月餅的數目顯著下跌，但根據推算，去年全港仍有超過187萬個月餅被丟棄，加起來的相等於78座大帽山的高度。"
3. I cannot find a corresponding press release or similar stuff in its homepage. My guess is : 0.81 * 2.3 million housholds = about 1.86 million mooncakes dumped.
This is a very dangerous practice, as the 300 households are far from being anything you can call representative - they are equivalent to only 0.02% of total households. Even the unemployment statistics that based on some 27000 households in Hong Kong are sometmes attacked by its lack of representativeness.
It is very expensive to collect samples. But it does not mean that you should blow up the figures in such imprudent or even reckless way just fot the sake of catching media attention.
4. So how reliable are the estimate?
No one can tell. But according to this source, Hong Kong sold about 3 million box mooncakes annually, most (perhaps 90%) are sold during autumn time and to Hong Kong people (over 90%). So roughly 80% of the 3 million boxes = 2.4 million boxes are consumed by HK.
The 1.86 million mooncakes are equivalent to some 0.5 million boxes (neglecting factors such as smaller/mini-mooncakes etc. for simplicity), or one-fourth of the mooncakes are dumped. And you think this figure for yourself.
I don't eat or buy any mooncakes. But my guess is, this is a figure on the high side.
Recent Comments