1
$\begingroup$

Autocorrelation is also known as serial correlation .

  • Why is the terminology serial used ? Is there anything unserial or disordered correlation ?

  • Also i see the word positive serial correlation frequently . But never had seen negative serial correlation . Is there anything "negative serial correlation" ? If so , can you please give me a real life example ?

Improve this question
$\endgroup$
1
  • 2
    $\begingroup$ Stock returns of very high frequency have negative serial correlation. The explanation is that the price is jumping between "bid" and "ask" randomly, and that is enough to create negative serial correlation. Some more examples of negative autocorrelation are mentioned here (this is the first thing I found, so not necessarily the best examples). $\endgroup$
    – Richard Hardy
    Commented Apr 17, 2015 at 11:11

1 Answer 1

Reset to default
1
$\begingroup$

alike Serial Dependance, if the value at some time t in the series is correlated on the some pair of value at another time s,say,serial correlation.as it is serially correlated. for this there is a term "no serial correlation"

Negative serial correlation implies that a positive error for one observation increases the chance of a negative error for another observation and a negative error for one observation increases the chances of a positive error for another.

you can test for type of auto-correlation by Durbin–Watson statistic

Real life examples

Improve this answer
$\endgroup$
5
  • $\begingroup$ What does the term "no serial correlation" imply ? Does it imply there is no correlation ? Or, does it imply there is disordered correlation ? $\endgroup$
    – time
    Commented Apr 17, 2015 at 11:34
  • $\begingroup$ "no serial correlation" means there is neither positive nor negative serial correlation. "no correlation" term used in linear regression it won't change by time. autocorrelation is time dependent, correlation not(at least independent for long time). I don't know if there is a word disordered correlation, because autocorrelations for non-stationary are usually not in order form. $\endgroup$
    – Hemant Rupani
    Commented Apr 17, 2015 at 11:46
  • $\begingroup$ If there is neither positive nor negative serial correlation , then it seems the pattern is random . And in regression it is the situation of "no correlation" . Then how does "no serial correlation" mean there is neither positive nor negative serial correlation ? What will be the pattern of it ? random ,like no correlation? $\endgroup$
    – time
    Commented Apr 17, 2015 at 12:16
  • $\begingroup$ Also in regression for checking the independence assumption , we usually plot residual versus time plot . If "no correlation" term used in linear regression won't change by time , then why should we plot "residual versus time" for checking the independence assumption . $\endgroup$
    – time
    Commented Apr 17, 2015 at 12:16
  • $\begingroup$ I am browsing [Lecture 8: Serial Correlation]. Have a look at it! google.co.in/url?q=http://www.columbia.edu/~so33/SusDev/… .... See pg.no.10 'no serial correlation' but you can't say 'no correlation' $\endgroup$
    – Hemant Rupani
    Commented Apr 17, 2015 at 12:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.