Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
17 Cards in this Set
- Front
- Back
|
What is a null hypothesis? |
The null hypothesis predicts the results that are anticipated by chance; that is, what we would expectdue to random occurrence. |
|
|
What is a alternative hypothesis? |
The alternative hypothesis predicts the results that occur due to particular bias; that is, becausesomething nonrandom has occurred. |
|
|
What is Significance and Confidence? |
Significance and confidence refer to the probability that a prediction about experimental results istrue and does not occur by chance. A significance level indicates your confidence in rejecting the nullhypothesis. How much doubt are you willing to live with? |
|
|
What are Levels of Significance? |
The criterion used for rejecting the null is indicated by the lowercase Greek symbol for alpha.Think of it as the level of doubt you are still willing to have when you assert your results are true |
|
|
What are common significance levels? |
A 0.01 level of significance corresponds to a 99% confidence level A 0.05 level of significance corresponds to a 95% confidence level A 0.10 level of significance corresponds to a 90% confidence level The lower the significance level and the greater your confidence (i.e. willingness to reject the null),the more the data is needed to support the alternate hypothesis |
|
|
What is the Test Statistic? |
The basis upon which decisions are made to either reject or fail to reject the null hypothesis Establishes whether the experimental outcome supporting the alternate hypothesis is statisticallysignificant |
|
|
What is theCritical Region? |
The area under the tails of a bell curve representing the difference required between the testelements with which one may conclude at a certain level of confidence that the results have notoccurred by chance alone. If the test statistic falls in the critical region the null hypothesis is rejected andthe alternative hypothesis is accepted |
|
|
What is the difference betweenOne-Tailed vs. Two-Tailed Tests? |
A one-tailed test is concerned with only one end of the sampling distribution. All testing for statisticalsignificance is done on this group of data in the “rejection” zone. A two-tailed test uses data from both edges of the bell curve (i.e., two rejection zones). Both one- and two-tailed tests help establish the alternate hypothesis – the experimental results areunlikely to have occurred because of chance. If the test statistic is beyond a critical value (i.e., theprobability that such results occur is less than or equal to the significance level), then the null hypothesisis rejected and the outcome is said to be statistically significant. |
|
|
What are Type I and Type II errors? |
a type I error is the incorrect rejection of a true null hypothesis (a "false positive"), while a type II error is incorrectly retaining a false null hypothesis (a "false negative"). A Type I error occurs when the team concludes that there is a significant difference when there really is not. A Type II error occurs when the team concludes that there is not a significant difference when there really is. |
|
|
How do you Resolve the problem of Type I and Type II Errors? |
Raise or lower the threshold. - Raising the threshold can result in type II errors (beta risk) - Lowering the threshold can result in type I errors (alpha risk) Reduce the variation in the signal. |
|
|
What are some guidelines to followFlooring Types when conducting a Hypothesis Testing? |
Start with an assertion (null hypothesis) Gather test results and make assertions based upon that Look at reality |
|
|
What is Beta risk? |
Raising the threshold too much, resulting in a type II error Type II errors are tied to the beta risk |
|
|
What is Alpha Risk? |
Lowering the threshold too much, resulting in a type I error |
|
|
What is the typical Value of Alpha? |
Alpha is often set to 0.05, or 5 percent, because: Most people believe that a 5 percent chance of making a type I error is a reasonable risk Most computer programs default to this when presenting statistical output Do not, however, accept this value simply because it is the industry standard. It is important to selectan appropriate alpha level for your particular application |
|
|
What is a typical Value of Beta? |
The beta risk is generally suggested to be 0.20, or 20 percent: Just like alpha, the reason we set betaat a particular level has to do with the consequences we are willing to accept if we make an error. |
|
|
What is Power? |
Power is the chance of finding the result or difference which we are examining: Power is equal to 100 percent minus the beta risk, or one minus beta Power depends on the level that we set beta It is preferable to have enough power to find whatever difference is meaningful to us |
|
|
What are some important elements to remember when making decisions using Hypothesis testing? |
The test stat of formulas. Some of them are simple, some of them are more complicated, but it isalways what we calculate. The critical value – or values if there is a two-tailed test – are the boundary lines to the rejection region,or the critical region. In hypothesis testing, the null hypothesis is always assumed to be true You never prove the null hypothesis |
