Time to Practice – Week Three

Complete both Part A and Part B below.

Part A

1.     What does the critical value represent? (2 points) The critical value represents a value an individual would assume the test at hand to produce if the null hypothesis were true. It is the statistic a person believes to find.

2.     Given the following information, would your decision be to reject or fail to reject the null hypothesis? Setting the level of significance at .05 for decision making, provide an explanation for your conclusion. (3 points)

a.     The null hypothesis that there is no relationship between the type of music a person listens to and his crime rate (p < .05).  Null hypothesis is rejected as p value is less than .05

b.     The null hypothesis that there is no relationship between the amount of coffee consumption and GPA (p = .62).  Since p=.62, which is more than .05 or greater we accept the null hypothesis.

c.     The null hypothesis that there is no relationship between the number of hours worked and level of job satisfaction (p = .51). since p=.51 greater than .05 accept null hypothesis.

3.     Why is it harder to find a significant outcome (all other things being equal) when the research hypothesis is being tested at the .01 rather than the .05 level of significance? (2 points) When the test is done at the .01 rather than .05 it produces a much small value set while .05 produces a larger one. This leads to the contradictory data which is quite larder in the research hypothesis and more likely.

Part B

Complete the following questions. Be specific and provide examples when relevant.

Cite any sources consistent with APA guidelines.

Question	Answer
What is a research question that you would like to answer? Write the null and research hypotheses. Would you use a one- or two-tailed test? Why? (3 points)	If the Health department is tasked with finding out the rate of lung cancer connected with smoking in the past year is it comparable to average rate past 50 years. Null hypothesis: No difference in last year and last 50 years. Research hypothesis: there is a difference in last year and last 50 years. Using a two-tailed test because of the research hypothesis would show that it was not equal
What do we mean when we say that a statistical result is significant? What is the difference between a statistically significant and a meaningful result? Why is statistical significance important?  (4 points)	Significant means that there is a rate of certain probability in something. Meaningful results state that the results are valid and statically significant which means a person rejects the null hypothesis. With there being some probable difference between the results we can conclude the null hypothesis is not acceptable and should be rejected.

From Salkind (2011). Copyright © 2012 SAGE. All Rights Reserved.

Reliability and Validity Matrix

For each of the tests of reliability and validity listed on the matrix, prepare a 50- to 70-word description of the test’s application. Describe what conditions these reliability types would be used for, as well as when they would be inappropriate. Then, for each test, prepare a 50- to 70-word description of the strengths and a 50- to 70-word description of the weaknesses.

Test of reliability	Application and appropriateness	Advantages	Disadvantages
Internal consistency	In internal consistency deals with assessing consistency of internal items or internal consistency of the items on a test.  This is used in order to estimate the reliability of items that are on the test without the need to develop alternate or additional forms of a test or having to give the test more than one time. Internal consistency employs a single measurement to estimate a test’s reliability.	One of the greatest advantages comes from only needing to administer the test one time and thus it provides a cost efficient and effective nature to the test taking process. Another advantage comes from internal consistency, which can be gained through a wide measurement of tests and through a variety of items.	One disadvantage is that high reliabilities with this method may actually be considered a weakness as it can indicate a redundancy in the test items and measures used. Another disadvantage comes from the results not being taken from the theory, which results in all variables being considered. The results are never perfect but rather have to be theorized to try to perfect the results.
Test or retest	With the use of this test a developer provides the test twice to an individual on different occasions. The developer compares the two scores in order to assure the testers haven’t changed from one test to the next. This method allows a way to estimate reliability, as there is a higher correlation between testers taking tests within shorter amounts of time between each test.	The advantage of this method is that it produces a lower score when it is compared with other methods. The time frame being shorter between first test and retest provide correlations between timeframes and lower scores being produced. Another advantage is that the method uses a single score or rater code.	Some disadvantages to the method is that it can be prone to weaknesses like from carryover effects attributing to or causing errors that are found in test scores. Timeframes can also affect scores produced by testers on the retake test. The method can also prove to be more costly since it requires the test be given multiple times and it can lack in reliability.
Parallel and alternate forms	Parallel test exist when all the additional or alternate forms are equal in means and variances of the observed test scores. The theory behind it is that means on parallel forms correlate equally to the true score of the test. Alternate forms are a method in which different test versions have been created to parallel the original test.	Advantages of these tests include that it can minimize effects that memory of content may have on administering previous forms of the test.  The tests may also provide stable scores and information in certain circumstances when measuring certain constructs.  Another advantages comes from being able to estimate reliability without having to develop another form of the test.	Disadvantages come from having to develop alternate forms of the test, which can be expensive as well as time consuming. Errors may result in variance or item sampling when trying to compute the alternate or parallel form of the reliability coefficient. Finally, the person taking the test may have performance issues, which are affected by a specific for of the test, done to the items included on that test.
Test of validity	Application and appropriateness	Strengths	Weaknesses
Content validity	Content validity is concerned with scrutinizing the content of the test as well as the extent to which the test measures represents all the facets for a given construct. A judgment is made of how adequately the test is able to sample behaviors that represent the universe of behavior in which the test was created for.	Advantages of the method involve defining as well as finding domain of items in which a creator of the test wants to measure. It helps to increase a test’s validity and leads to content validity. Items on the test go with things wanting to be measured without deviating from items that want to be measured.	The disadvantages of this method are that can require experts to design, develop, and evaluate the test and scores. This makes the test both expensive as well as time consuming.  Furthermore, the test may have a need to cover a variety of items and information that can be long for either administering or taking the test.
Criterion related	Criterion related is used in order to demonstrate accuracy of a procedure or measure by comparing the procedure with another one to demonstrate it to be valid. It is used as a judgment of the adequacy of the test to basically predict or infer a person’s score on a given subject.	One advantage of the method is that involves use of two estimators to demonstrate validity of a test that has been given. This method works for academic use in being able to predict scores. Additionally it can work well in the determination of how certain things like traits develop overtime.	Disadvantages associated with the method include accuracy. Individuals change overtime which can lead to problems with accuracy or reliability with the tests. Another disadvantage is in the change to academics itself, individuals learn one thing and it is changed on the test or it needs to be changed on the test making it costly to do so.
Construct	Construct deals with a test or an experiment being able to measure what it claims to do.  It refers to the operational definition of a variable actually reflecting the theoretical meaning of a certain concept.  This method requires judgment about appropriateness of the scores drawn from the inferences in a person’s standings.	One advantages of this method include that domain of an item or the behavior wanting to be measured is obtained. The domain behavior is kept in mind as the items on the experiment or test are reviewed so all items can show a relationship with behaviors wishing to be measured.	Disadvantages of the method require experts in the areas of the test or experiments to help with the tests. Additionally experts are also needed in the use of the experiments and evaluation of tests that makes the test more time consuming as well as being more costly to create, perform, as well as evaluate. Finally, the mere development of the test is not easy either.

Basic Concepts in Statistics

Differentiate between descriptive and inferential statistics. What information do they provide? What are their similarities and differences? Your answer should be 250 to 400 words.

Descriptive statistics are values that describe the characteristics of a sample or a population. Descriptive statistics allows a large amount of information or data to be used in order to come up with a summary of the data (Salkind, 2014). 

. For instance if you had a list of students at a school from the same grade and listed all their grades in differing subjects you could use descriptive statistics in order to find the average or mode of the kids grades. Descriptive statistics is able to take either a large sample or a population and turn the list of information, data set, into a summary that researchers can then use in different capacities such as predicting the grades of students across an age group.  If the researcher wanted to see what subject children excel more in then through the use of the grades and the differing subjects they would be able to find that information. Inferential statistics is different as it is the tools that are used in order to infer the results based on a sample to the population (Salkind, 2014). Inferential sometimes follows descriptive statistics as the next step in the process taking the small group that may have been used and applying the information to a larger group or population.  The inferential statistics allow researchers to make a kind of guess into how the summary that was created may apply to a larger population.  If researchers took a sample of a few different students from several schools in order to find grades in subjects they could then use the results to infer what the whole population may look like.

What is a population? What is a sample? How are they similar and how are they different? When would you use one or the other? Your answer should be 250 to 400 words.

A population according to Salkind (2014).  Is all the possible subjects or cases of interest.  So if a person wanted to know how good a soccer team was then they would use all individuals on the soccer team as the population when doing calculations.  While sample is a subset of a population (Salkind, 2014). It is much like its name describes and only a small portion of the total population. So in the soccer team you would take 10 players and this would be the sample. Both are representatives of the data that a person is trying to produce with one entailing all the subjects involved and the other only being made up of a portion.  When a population is relatively small such as the example of the soccer team then it is possible to use the entire population in the calculations being done. However when the population is larger such as looking at teen drug use then the population would include a size so big that it would be relatively impossible to look at all teens in the world. It would take too much time any money in order to do this. In this case a sample is pulled in order to represent the entire population while still getting the insight that a behavioral scientist is looking for. That way the scientist is able to pull a relatively small amount in comparison to the entire globe and is able to make inferences on who the results apply to the entire population of teens.

Salkind, Neil, J.(2014). Statistics for People Who Think They Hate Statistics. 5th Edition. Sage Publications.