Statistics Help @ Talk Stats Forum - Statistics
http://www.talkstats.com/
Statistics course and homework discussion. Elementary statistics.enTue, 13 Oct 2015 12:48:08 GMTvBulletin60http://www.talkstats.com/images/misc/rss.pngStatistics Help @ Talk Stats Forum - Statistics
http://www.talkstats.com/
Sample Size Estimation for Dependent or Independent Groups?
http://www.talkstats.com/showthread.php/62685-Sample-Size-Estimation-for-Dependent-or-Independent-Groups?goto=newpost
Mon, 12 Oct 2015 18:54:34 GMTHey everyone,
I am new to this so please bear with me. I have been discussing a question with my fellow peers regarding sample size estimation....Hey everyone,

I am new to this so please bear with me. I have been discussing a question with my fellow peers regarding sample size estimation. The question is as follows;

Suppose we wish to design a study to compare two antihypertensive medications. Two outcome variables will be considered: systolic and diastolic blood pressure.

a) How many subjects would be required to detect a difference of 15 mmHg in mean systolic blood pressure between groups with 80% power and alpha=0.05 (two-sided)? Assume that the standard deviation in systolic blood pressure is 30 mmHg.

b) How many subjects would be required to detect a difference of 6 mmHg in mean diastolic blood pressure between groups with 80% power and alpha=0.05 (two-sided)? Assume that the standard deviation in systolic blood pressure is 12 mmHg.

I've been discussing my answer with a few people and they are getting different answers. I've been using sample size estimation for two dependent groups, yet they are using sample size estimation for two dependent groups (with effect size). Would anyone be able to explain which once I need to use and why? I ended up getting 32 subjects for a) and b) using the two dependent groups equation while my peers ended up with 63 subjects for a) and b) using two independent groups equation.

Thanks for your help all!
]]>StatisticsBananashttp://www.talkstats.com/showthread.php/62685-Sample-Size-Estimation-for-Dependent-or-Independent-GroupsWhich statistical test?
http://www.talkstats.com/showthread.php/62683-Which-statistical-test?goto=newpost
Mon, 12 Oct 2015 16:17:45 GMTI am returning to statistics after a gap of a few years, I was never confident before and seem to be really struggling now! I was wondering if...I am returning to statistics after a gap of a few years, I was never confident before and seem to be really struggling now! I was wondering if anybody on here could kindly help me with a couple of problems.

There are two questions that I am trying to answer:

1. "Test the effects of two different diets on the blood plasma calcium concentration in 4-week old rats."

For this would I be correct to use a t-test for unmatched samples?

And...

2. "Determine the effects on the growth of tomato plants of 4 different levels of nitrate and 4 different levels of phosphate added to the soil. What test would you use to determine whether there were any significant effects of phosphate and/or nitrate on the growth rates of the tomato plants?"

We must look at each treatment individually. This is the one I'm finding really tricky but I thought that maybe a regression analysis and pearsons correlation coefficient might be correct?

Any help would be hugely appreciated!
]]>Statisticsnewtostats2http://www.talkstats.com/showthread.php/62683-Which-statistical-testCovariance same as standard score and raw score?
http://www.talkstats.com/showthread.php/62678-Covariance-same-as-standard-score-and-raw-score?goto=newpost
Sun, 11 Oct 2015 20:24:31 GMTI have a problem in which we have to compute the correlation coefficient, using the standard score formula, covariance formula, and the raw score...I have a problem in which we have to compute the correlation coefficient, using the standard score formula, covariance formula, and the raw score formula. Using the raw score and standard score formulas, I got very similar answers, but my covariance formula is giving me a very different answer. Are all 3 supposed to be similar??
]]>StatisticsJWadehttp://www.talkstats.com/showthread.php/62678-Covariance-same-as-standard-score-and-raw-score