e

主頁

書

音樂

介紹

攝影

心理學

my blog

You might heard the joke about psychology being telling you something that you already knew in words that you don't understand ^_^

But let's try these simple experiments and tests and see if you can think of a common-sense explanation, and then compare with the ones the psychologists came up with ^_^

Stroop Task

This is a classic test devised by Stroop in the 1930s, and which is still widely used in many different fields. If your first language is Chinese, like me, then it would be better to use Chinese characters in this experiment, but let's try and see how it goes... (It's even better to use a stopwatch to time the time.)

First, you need to read the words as fast as possible.
Red	Green	Blue	Yellow
Blue	Green	Yellow	Red
Yellow	Blue	Red	Green
Green	Red	Yellow	Blue
Again, you need to read the words as fast as possible.
Red	Green	Blue	Orange
Blue	Green	Orange	Red
Orange	Blue	Red	Green
Green	Red	Orange	Blue
Next, you need to name the colour... in English.
XXX	XXX	XXX	XXX
XXX	XXX	XXX	XXX
XXX	XXX	XXX	XXX
XXX	XXX	XXX	XXX
Finally, name the colour of the words, not reading the word.
Red	Green	Blue	Orange
Blue	Green	Orange	Red
Orange	Blue	Red	Green
Green	Red	Orange	Blue

Your should find that the first two are the fastest, then the third, and the last slowest. It would be clearer if the words and in Chinese and you need to name them and the colour in Chinese.

But it neatly illustrates the mechanism behind the results. Basically we are very used to reading out words, and less used to naming something's colour. That's why the reading of words are very much an automated process. And when we see a word, particularly in our first language, we have a natural tendency to read it.

In the fourth task, we need to surpress the tendency to read out the words, in favour of a less automated process. There is a suggestion of a competition between the two process, and a well-practiced process would have an initial advantage; and we need time to surpass it until the other process "wins".

Having this initial results, we can draw further hypotheses about the processes. For example, we would expect people who are used to naming colours would should less effect, or even a reversed effect; and also for people working on different languages.

We can then design experiments to test whether these hypotheses are right. That's what psychology is about... Hypotheses Testing.

您可能有聽過這個笑話：「心理學就是用您聽不懂的術語來說您本來就知道的東西。」^_^

或許您也這樣想吧。這裡我介紹幾個簡單的實驗，您可以試試看用常識來解釋，然後和所謂心理學家的解釋比較看看 ^_^

Stroop Task

這是Stroop在1930年代發明的一個經典實驗——而它和它的變種到現在仍被廣泛使用。 它基本上是一個計算回應速度（Reaction Time）的實驗，所以如果您可以計算您自己的速度就更好了。

首先，請以最快的速度讀出以下的字。
紅	綠	藍	黃
藍	綠	黃	紅
黃	藍	紅	綠
綠	紅	黃	藍
同樣，請以最快的速度讀出以下的字。
藍	綠	橙	紅
紅	綠	藍	橙
綠	紅	橙	藍
橙	藍	紅	綠
這次，請講出每組「xxx」的顏色。
XXX	XXX	XXX	XXX
XXX	XXX	XXX	XXX
XXX	XXX	XXX	XXX
XXX	XXX	XXX	XXX
最後，請講出下面每個字的顏色。
橙	藍	紅	綠
紅	綠	藍	橙
綠	紅	橙	藍
藍	綠	橙	紅

一般來說，第一和第二個部份應該是最快的；其次是第三部份；而第四部份最慢。

跟據心理學家的研究，這實驗反映出我們大腦處理這些資料的機制。 基本上，我們從小就習慣閱讀，而比對上不太習慣說出物件的顏色。 因此，朗讀對我們來說幾乎是一個自動的過程。 而當我們見到一個字的——尤其是那是我們的母語，我們很自然的就會把它讀出來。

在第四部份，我們需要「壓抑」我們朗讀的傾向，而選擇一個沒那麼自由的過程。 一般來說，兩個過程之間會有一種競爭，而一個自動的、或常用的過程會有一點優勢。 而我們就需要時間去壓抑這個過程，自至另外的過程「獲勝」為止。

根據以上的結論，我們可以想出其他的假說（hypothesis）。 例如，如果以上的結論正確，那某些慣常處理顏色的人應該會有中性或甚至相反於一般人的Reaction Time； 而中文不是母語的人也有不同。

當我們有這些假說後，我們便可以設計新的實驗去驗証這些新的假說。 而這「Hypotheses Testing」正是心理學的發展方法。

Visual Search

This is a very general test, and the basic idea is to see what factors affect the time one needs to detect a target in a field. Let's try to find a blue H in the following displays. (It could be absent.)

	H	H	H
H	H		H
	H	H		H
		H	H
H	H	H

		H	H
H	H		H
H	H	H
	H	H	H
	H	H		H

		H	E
E	E		H
H	H	E
	E	H	H	E
	H	H		E

H	H	E
E	E		H
	H	H		E
	E	E	H	E
		H	E

The usual result is that the first two is easier than the last two. The idea is that the blue H in the first display would "pop-out", and in the second, nothing pops out, so in both display we can quickly determine whether the target is present or absent. These are called the "single feature search". The single feature here is a colour (blue), but it can also be size, shape, movement, etc.

On the other hand, in the last two displays, there are two features, "H" and "blue". It is because by using only one criteria, say blue, we cannot eliminate the blue "Es"; and using the criteria "H", we cannot eliminate the red "Hs". That's why we need to perform a linear search, i.e. checking the shapes one by one.

The time we need for a single feature search are not related to the number of distractors (like the red Hs), because the target would just pop-out. For more than one feature, however, we need more time when there are more distractors, because we check them one by one. Also, the time we need for the absent trials are on average double that for the present trails, because in the present trails we would on average found the target after we went through half the display.

Last updated: 1 Jul 2004

Visual Search

This is a very general test, and the basic idea is to see what factors affect the time one needs to detect a target in a field. Let's try to find a blue H in the following displays. (It could be absent.)

	H	H	H
H	H		H
	H	H		H
		H	H
H	H	H

		H	H
H	H		H
H	H	H
	H	H	H
	H	H		H

		H	E
E	E		H
H	H	E
	E	H	H	E
	H	H		E

H	H	E
E	E		H
	H	H		E
	E	E	H	E
		H	E

The usual result is that the first two is easier than the last two. The idea is that the blue H in the first display would "pop-out", and in the second, nothing pops out, so in both display we can quickly determine whether the target is present or absent. These are called the "single feature search". The single feature here is a colour (blue), but it can also be size, shape, movement, etc.

On the other hand, in the last two displays, there are two features, "H" and "blue". It is because by using only one criteria, say blue, we cannot eliminate the blue "Es"; and using the criteria "H", we cannot eliminate the red "Hs". That's why we need to perform a linear search, i.e. checking the shapes one by one.

The time we need for a single feature search are not related to the number of distractors (like the red Hs), because the target would just pop-out. For more than one feature, however, we need more time when there are more distractors, because we check them one by one. Also, the time we need for the absent trials are on average double that for the present trails, because in the present trails we would on average found the target after we went through half the display.

Last updated: 10 Jul 2004