95% Confidence Interval

huitzilopochtli · 14-10-2012 9:26pm #1

Having a spot of bother with this. I understand it to be a range in which you can be 95% confident your sample will fall. Pretty sure that's right. My problem is that I don't understand the figure I got for it.

Here's a rough idea of what I'm doing:
I took 15 samples in a survey to get a mean of 1.73 organisms per sample.
Standard deviation worked out to be 1.53 and standard error 0.40. According to my handout 95% confidence interval is calculated by s.e. multiplied by a coefficient (2.145 on my table as I have 14 degrees of freedom - as I took 15 samples) and it comes out as... 0.86. What the heck does that mean?

My (poorly) educated guesses:
95% of the time I can be confident of finding 0.86 organisms in my sample.
95% of the time I can be confident of finding 1.73+/-0.86 organisms in my sample.
I mucked up somewhere in calculating the confidence interval.

Can anyone help me here?

Improbable · 20-10-2012 8:27pm

Maybe this will help:

http://www.statisticshowto.com/articles/how-to-find-a-confidence-interval-for-a-sample/

2Scoops · 20-10-2012 11:14pm

You're trying, presumably, to generalize the results from your sample to the mean of the population from which the sample is drawn. The 95% CI indicates that there is a 95% chance that the 'true' population mean falls within the range 1.73+/-0.86.

beardedmaster · 21-10-2012 12:51pm

Biostats in UCD by any chance?

huitzilopochtli · 21-10-2012 2:23pm

Thanks 2Scoops. Sorted now.

@beardymaster - no, TCD Zoology.

beardedmaster · 21-10-2012 3:23pm

Most people doing Science in UCD have to do a module on Biostatistics at some point, and we're covering that exact topic at the moment, so was just checking.

huitzilopochtli · 21-10-2012 3:47pm

If I had statistics lectures I wouldn't need to ask. We don't get anything till next term.

95% Confidence Interval

Comments