Timeline for How to split a numeric variable into a binary low-high variable
Current License: CC BY-SA 3.0
15 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 2, 2011 at 7:55 | comment | added | Jeromy Anglim | @Orp I've altered this question in line with the comments with the aim of making it clearer for others in the future. Feel free to modify if I have misconstrued what you are asking. | |
| May 2, 2011 at 7:54 | history | edited | Jeromy Anglim | CC BY-SA 3.0 | I've altered this question in line with the comments with the aim of making it clearer for others. Feel free to modify if I have misconstrued.; added 20 characters in body |
| May 2, 2011 at 4:47 | answer | added | Jeromy Anglim | timeline score: 8 | |
| May 2, 2011 at 3:11 | history | tweeted | twitter.com/#!/StackStats/status/64889674077970433 | ||
| S May 1, 2011 at 23:36 | history | suggested | rolando2 | CC BY-SA 3.0 | These edits are the only way the post would make sense to me. |
| May 1, 2011 at 23:17 | answer | added | rolando2 | timeline score: 6 | |
| May 1, 2011 at 23:08 | review | Suggested edits | |||
| S May 1, 2011 at 23:36 | |||||
| May 1, 2011 at 22:13 | comment | added | Avery Richardson | I'm not sure that I understand the variables in your study. It seems that you have one continuous independent variable and at least one continuous response variable. You say that you would like to cluster the independent variable. Why not just use it as a continuous variable without clustering? I think this is like what KennyPeanuts asked. | |
| May 1, 2011 at 20:33 | comment | added | Orp | KennyPeanuts - the data come from a sample of 15 individuals, so I do not assume normality. My aim is to compare the scores of subjects that exhibit high levels of the observed behavior with subjects that show low levels of the behavior on a different score. Therefore I would like to cluster all of them into two distinct groups based on their level of expression (high vs. low). Thank you. | |
| May 1, 2011 at 20:08 | comment | added | Orp | Thanks Jeromy, I am not sure I understand your question. The behavior scores range from 0 occurrences per minute to 8 per minute on a continuous scale. | |
| May 1, 2011 at 15:20 | comment | added | DQdlM | @Orp there seem to be a lot of different ways to approach this... some more information is needed to best develop an answer. How are the data distributed? Do you have an a priori reason to think there are 2 groups and only 2 groups? etc... | |
| May 1, 2011 at 14:47 | comment | added | Jeromy Anglim | @Orp Is the behaviour measured on a single variable or multiple variables? | |
| May 1, 2011 at 13:46 | comment | added | Orp | The levels of the behavior are the frequencies (occurrence of behavior/minute) that were exhibited by the individuals. | |
| May 1, 2011 at 11:35 | comment | added | user88 | It seems like some clustering, but can you elaborate a bit more? What do you mean by "levels of that behavior"? | |
| May 1, 2011 at 11:20 | history | asked | Orp | CC BY-SA 3.0 |