In statistics, we have countless of tests to perform on the data.
But what kind of test to use?
What kind of test for what kind of data?
Under what condition, the test can be performed?
This book will serves as a guide book for young statistician and the user of statistics.
User is, of course, required to know:
1. What kind of data that they are using- linear or circular data?
2. Amount of sample for testing- 1 sample, 2 samples or many samples?
3. Using parametric test or non-parametric test?
Most of the tests will be using hypothesis testing to test mean, variance, median, randomness or distribution. Last but not least, user is expected to know whatever to use one or two-tailed test before the test is even conducted. This is crucial for making a right decision.
I randomly pick one of the pages to serve as a sample. This test is used to investigate the difference of the means for two populations. In this example, the author wish to find: Is there a difference between average of two pack of potato crisps (in term of weights)?
Well, good thing you don't have to read through all the statistics books just to find the test you want. Honestly, this is one of the good statistics book because the author gives brief objective of the test at the beginning, the limitations of the test, methods (including test statistic and statistics table), example that related to real life application and lastly numerical calculation at the end of the test. This book has few setbacks such as the example is made to be as simple as possible and the full data is not there, but this will not prevent user from understand the test.
p/s: I think as a statistician, one must know what kind of test that need to perform on the data, how to derive a simple test statistic like the example above and most importantly making a right decision.
But what kind of test to use?
What kind of test for what kind of data?
Under what condition, the test can be performed?
This book will serves as a guide book for young statistician and the user of statistics.
User is, of course, required to know:
1. What kind of data that they are using- linear or circular data?
2. Amount of sample for testing- 1 sample, 2 samples or many samples?
3. Using parametric test or non-parametric test?
Most of the tests will be using hypothesis testing to test mean, variance, median, randomness or distribution. Last but not least, user is expected to know whatever to use one or two-tailed test before the test is even conducted. This is crucial for making a right decision.
I randomly pick one of the pages to serve as a sample. This test is used to investigate the difference of the means for two populations. In this example, the author wish to find: Is there a difference between average of two pack of potato crisps (in term of weights)?
Well, good thing you don't have to read through all the statistics books just to find the test you want. Honestly, this is one of the good statistics book because the author gives brief objective of the test at the beginning, the limitations of the test, methods (including test statistic and statistics table), example that related to real life application and lastly numerical calculation at the end of the test. This book has few setbacks such as the example is made to be as simple as possible and the full data is not there, but this will not prevent user from understand the test.
p/s: I think as a statistician, one must know what kind of test that need to perform on the data, how to derive a simple test statistic like the example above and most importantly making a right decision.
No comments:
Post a Comment