h**l 发帖数: 4883 | 1 我朋友拿来一组数据让我分析。由于他不懂统计,采集pre和 post的数据的时候,
randomize了采集数据的对象。也就是说pretreatment 和posttreatment没发match上。
但他又想测两个factor (related and unrelated).
这样的话,他有
Pre post
Related a b
Unrelated c d
a, b,c,d分别是四组数据。
他想测(b-a) 和(d-c) 之间有没有区别。该怎么做,因为没法match对象,我也想不出
该用什么test... | D******n 发帖数: 2836 | 2 why testing that? whats the purpose?
what are related and unrealted ?
btw, they are two levels of one factor in stead of 2 factors.
you have 2 factors here, treatment and status(rel/unrel)
【在 h**l 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 我朋友拿来一组数据让我分析。由于他不懂统计,采集pre和 post的数据的时候, : randomize了采集数据的对象。也就是说pretreatment 和posttreatment没发match上。 : 但他又想测两个factor (related and unrelated). : 这样的话,他有 : Pre post : Related a b : Unrelated c d : a, b,c,d分别是四组数据。 : 他想测(b-a) 和(d-c) 之间有没有区别。该怎么做,因为没法match对象,我也想不出 : 该用什么test...
| h**l 发帖数: 4883 | 3 testing scores, so it's ratio data and should useparametric test.
Related and unrelated are just two factors.
I've tested using ANOVA already and there are statistical differences
between a, b, c and d. Posthoc tests showed there were diferences between a
and b, a and c, a and d, b and c, b and d, c and d....
But he wants to test if there is any difference between (a-b) and (c-d).
【在 D******n 的大作中提到】![](/moin_static193/solenoid/img/up.png) : why testing that? whats the purpose? : what are related and unrealted ? : btw, they are two levels of one factor in stead of 2 factors. : you have 2 factors here, treatment and status(rel/unrel)
| h**l 发帖数: 4883 | 4 I used the dummy variables for that.
【在 D******n 的大作中提到】![](/moin_static193/solenoid/img/up.png) : why testing that? whats the purpose? : what are related and unrealted ? : btw, they are two levels of one factor in stead of 2 factors. : you have 2 factors here, treatment and status(rel/unrel)
| A*******s 发帖数: 3942 | 5 don't understand your question. But if you lost ID information in Pre-Post
comparison data, the data are useless.
【在 h**l 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 我朋友拿来一组数据让我分析。由于他不懂统计,采集pre和 post的数据的时候, : randomize了采集数据的对象。也就是说pretreatment 和posttreatment没发match上。 : 但他又想测两个factor (related and unrelated). : 这样的话,他有 : Pre post : Related a b : Unrelated c d : a, b,c,d分别是四组数据。 : 他想测(b-a) 和(d-c) 之间有没有区别。该怎么做,因为没法match对象,我也想不出 : 该用什么test...
| D******n 发帖数: 2836 | 6 use contrast?
a
【在 h**l 的大作中提到】![](/moin_static193/solenoid/img/up.png) : testing scores, so it's ratio data and should useparametric test. : Related and unrelated are just two factors. : I've tested using ANOVA already and there are statistical differences : between a, b, c and d. Posthoc tests showed there were diferences between a : and b, a and c, a and d, b and c, b and d, c and d.... : But he wants to test if there is any difference between (a-b) and (c-d).
| h**l 发帖数: 4883 | 7 I feel the data is useless too. But now we are trying to analyze it by
treating the data as independent variables. He spent a lot of time
collecting these data so I don't want to disappoint him.
【在 A*******s 的大作中提到】![](/moin_static193/solenoid/img/up.png) : don't understand your question. But if you lost ID information in Pre-Post : comparison data, the data are useless.
| h**l 发帖数: 4883 | 8 how?
【在 D******n 的大作中提到】![](/moin_static193/solenoid/img/up.png) : use contrast? : : a
|
|