Page 1 of 1: Q: Biostatistics: Which test to test hypothesis?

tallpox · Credits: 12032 My Scrapbook

. An investigator wants to study the association between
maternal intake of iron supplements (Yes/ No) and birth weights
(in gms) of newborn babies. He collects relevant data from 100
pregnant women and their newborns. What statistical test of
hypothesis would you advise for the investigator in this situation
?
1. Chi-Square test.
2. Unpaired or independent t-test.
3. Analysis of Variance.
4. Paired t-test.
Ans 2

decembermist · Credits: 41021 My Scrapbook

neelanjana · Credits: 2086 My Scrapbook

ans is unpaired t test since here we hav 2 samples of which one is a quantitative data (birth wt).
chi square test is used for qualitative data which is in the form of proportions.
ANOVA used when more than one comparison is to be made.(t test can make only one comparison).
paired t test used 2 compare 2 paired smples.(rfrnce mahajan)

gururulz · Credits: 132 My Scrapbook

ya i guess its T test but.,....why not chi square test???...i mean if weights r in the form of nominal data,,......then even chi square mebbe a possiblity

neone??

sweetea · Credits: 110 My Scrapbook

i think it is independent /unpaired T Test. where is ANOVA used?

abuye · Credits: 121 My Scrapbook

i agree with t test. chi square test are meant for nonparametric stats

shuvo · Credits: 4585 My Scrapbook My Reading List 11 Books

this will be independent sample t-test. cannot be chi-square as weight is taken in grams which is a continuous quantitative variable.

balaji_doc · Credits: 327 My Scrapbook

hi,
chi square test is for qualitative data i.e., non numerical or some category. This question and a similar confusing one was asked in may and november 2003,.
sinple tecnique is that whenver ur able to put the question in YES OR NO FORMAT its basically chi square test u have to use. Eg. association between iron intake and low birth weight. BIRTH WEIGHT cannot be put in a YES OR NO format. But LOW BIRTH WEIGHT can be since PROPORTION OF BABIES WEIGHING LESS THAN 2500g IN ONE CATEGORY AND BABIES WEIGHING ABOVE 2500g IN ANOTHER CATEGORY. BUT IT CANNOT BE APPLIED FOR BIRTH WEIGHT. SINCE IT A CONTINUOUS VARIABLE AND IS QUANTITATIVE AND IS DISTRIBUTED IN ALL PEOPLE. BUT LOW BIRTH WEIGHT IS QUALITATIVE ie., A CATEGORY WHICH IS YES IF BABIES WEIGH <2500g AND NO IF WEIGH ABOVE 2500g. SO HERE CHI SQUARE TEST ONLY SHOULD BE USED. I AM SURE. FOR FURTHER CLARIFICATION REFER MUDIT KHANNA 2006-2002 VOLUME.
I WAS TOO AFRAID OF Biostatistics . BUT NMS Biostatistics CHAPTER SIMPLIFIES EVERYTHING. U CAN ALSO REFER ACROSS PSM AND ASHISH GUPTA AND MUDIT KHANNA. IT STRENGTHENS UR CONCEPTS..

jabby · Credits: 3703 My Scrapbook

unpaired or independent T test i think is correct.chi square deals nominal data,any number of groups.
ANVOA simlar to Z/T test but it compares mean value from three or more groups simultaneously on one or more factors.
paired t test deals e matched study if each person in one gr is matched with a person in the second,applies to before and after measures and linked data.

aquarian · Credits: 257 My Scrapbook

unpaired t / student's t test-- variability b/w mean of a populationas a whole n tht of a sample drawn from it or 2 samples drawn from 2 diff pop.

paired t test -- significance of variablility of values of a sample of pop before n after exposure to a particular experience

chi-square test-- siginificance of assoc b/w any 2 QUALITATIVE data

correlation test -- strength of assoc b/w 2 QUANTITATIVE data

		Sign In New User? Sign Up Sign in to access your control panel and messenger!
				TechZone \| SpiderNevi \| HowTo? \| Scrapbook!