My Math Forum (intro to stats) How to Test a Research Question? p-value or traditional method

 Probability and Statistics Basic Probability and Statistics Math Forum

 December 15th, 2016, 08:22 PM #1 Newbie   Joined: Dec 2016 From: USA Posts: 1 Thanks: 0 The assignment is: "Use the five-step process for testing a research question. You can use the traditional method or the p-value method." I literally don't know where to begin. Basically, in my class the prof. put up a survey and the class answered the survey, and that is our data. (the survey was converted to numbers in an excel file) (the numbers correspond to an answer choice on the survey, like a key. ex/1=dog, 2=cat) We then have to pick two questions as testable variables and somehow test it. I think for a correlation, or a p-value maybe. The two questions on the survey I chose are this: 1) What is your gender? 2) Do you prefer cats or dogs? My claim is this: I believe the majority of females will prefer cats, and the majority of males will prefer dogs. How would I test the hypothesis between these two variables? That is, how can I either reject the null hypothesis or fail to reject it? Thanks, Clara Last edited by skipjack; December 16th, 2016 at 02:53 PM.
 December 16th, 2016, 10:51 AM #2 Senior Member     Joined: Sep 2015 From: USA Posts: 2,530 Thanks: 1390 5 seconds worth of google https://learn.bu.edu/bbcswebdav/pid-...fivesteps.html sheesh
December 17th, 2016, 06:49 PM   #3
Senior Member

Joined: Dec 2012
From: Hong Kong

Posts: 853
Thanks: 311

Math Focus: Stochastic processes, statistical inference, data mining, computational linguistics
Quote:
 Originally Posted by claraflower The assignment is: "Use the five-step process for testing a research question. You can use the traditional method or the p-value method." I literally don't know where to begin. Basically, in my class the prof. put up a survey and the class answered the survey, and that is our data. (the survey was converted to numbers in an excel file) (the numbers correspond to an answer choice on the survey, like a key. ex/1=dog, 2=cat) We then have to pick two questions as testable variables and somehow test it. I think for a correlation, or a p-value maybe. The two questions on the survey I chose are this: 1) What is your gender? 2) Do you prefer cats or dogs? My claim is this: I believe the majority of females will prefer cats, and the majority of males will prefer dogs. How would I test the hypothesis between these two variables? That is, how can I either reject the null hypothesis or fail to reject it? Thanks, Clara
You can try Fisher's exact test (which is more complicated, though more accurate and probably easier to do with the p-value method) or Pearson's chi-squared test (which is less accurate and very easy to do with the traditional method).

 Tags intro, method, pvalue, question, research, stats, test, traditional

 Thread Tools Display Modes Linear Mode

 Similar Threads Thread Thread Starter Forum Replies Last Post groverdot Advanced Statistics 2 December 5th, 2013 04:58 PM statsman1 Advanced Statistics 0 October 26th, 2012 03:14 AM Tharsos Advanced Statistics 0 July 2nd, 2012 07:25 PM gigsaw Advanced Statistics 1 March 26th, 2012 01:59 AM ohsaki Advanced Statistics 1 April 12th, 2010 08:45 AM

 Contact - Home - Forums - Cryptocurrency Forum - Top