Metrics-Final

If X was -1 in the regression equation Y=3+2X, the predicted value for Y would be:

1

What is the median of the vector- 11,12,13,14,15,16,17

14

What is the median of the vector-23,25,27,2000

26

If you have a z-value of 0.71 how much of the area under the curve is between the mean and the z-value

26.11%

What is the mode-3,4,6,3,4,3,8,4,9

3,4

What is the mode of the vector 4,5,4,6,7,9,4,8

4

On the last quiz you got a 64, and the class average was 85 with a SD of 10. Is your score significantly different from the rest of the class? Assume that you want to be 95% certain (p<.05) A) Your score is significantly lower than expected OR B) Your score is low, but not significantly higher than expected

A

A dictionary-like statement that describes a construct is the: A) Nomological Network B) Construct Domain C) Internal Consistency

B

A manager of one of you company's stores calls you up to b/c her store got a 2.3 on the last engagement survey. She wants to know if her store's results significantly differ from the other store. You look up the mean and SD and note that means engagement across all stores is 3.1 and the SD is 0.5. Do you tell the manager A-your score is significantly lower OR B-Your score is low, but not significantly lower

B

While working to define a measure of cohesion, your team identifies the following items: (Which item appears to represent a contamination error?) A) My closest relationships are with my colleagues B)My supervisors are supportive C)My colleagues would help me with a personal problem D) I choose to spend time with my colleagues after work

B

List Google's people analytics group four levels of sophisticated in HR Analytics

Counting, clever counting, insight and influence

Assume you have a normal distribution if you take 2 SD, how much does it cover?

Covers 95%

....statistics are used to describe or summarize data; ....statistics are used to make inferences about a population(Fill in the blank)

Descriptive and Inferential

"a" represents the slope in the regression equation "y=a+bX (T/F)

False

A best practice in survey design for job-engagement surveys is to put questions about respondent demographics in the first part of the survey(T/F)

False

A cronbach alpha coefficient alpha value of 0.60 or greater is generally considered acceptable reliability(T/F)

False

A dependent variable, often denoted as Y, is assumed to influence another variable(T/F)

False

A good survey will put boxes and labels around certain sets of items to keep them distinct from other items(T/F)

False

A skewed distribution is symmetrical(T/F)

False

An independent variable is assumed to be influence by another variable. That is the independent variable is assumed to be the outcome(T/F)

False

Good survey items ask about thoughts and motives rather than about observable behavior(T/F)

False

If a decrease in X is associated with a decrease in Y, the correlation coefficient would be negative(T/F)

False

If a scatter plot were curvilinear, it would likely produce a correlation close to 0.5(T/F)

False

If turnover is highly correlated with firm performance such that firms with low turnover have high performance, then that mean turnover is the factor causing performance(T/F)

False

If you follow the construct validation steps you can decisively conclude that your construct has been "validated". That is, you can be completely confident that you have a validated construct (T/F)

False

If you want to contrast intrinsic motivation scores for a group of exempt versus a group of non-exempt employees you would probably use a matched/related/dependent t-test (T/F)

False

If your boss wanted to know if performance ratings (collected on an interval scale from 1-100) differed for males and females, you would be correct to suggest a chi-square test(T/F)

False

In a graph, the Y-axis is the horizontal line(T/F)

False

In a regression equation, the x-variable is the dependent variable(T/F)

False

Inferential statistics are used to make statements about a sample when we know the characteristics of a population(T/F)

False

It is often practical and feasible to collect data from an entire population(T/F)

False

Leaders of HR profession are aware about and evaluate HR(T/R)

False

Leadership support is a common assumed driver of employee attitudes such as commitment(T/F)

False

One of the main reasons why it's useful to know about the chi-square test because the test is useful for analyzing interval data such as means form engagement surveys(T/F)

False

One way to improve survey items is to include two topics in one question such as "my manager is good at hiring staff and setting fair compensation"(T/F)

False

The formula below is the formula for the median(Sigma X/N)

False

The function EXCEL function VAR.S estimates the variance for a population(T/F)

False

The one exception to the rule "correlation does not equal causation" is when you observe a correlation that is so significant, the t-value is greater than 10 and the p-value is less than .001(T/F)

False

The regression line is the line that passes through the X,Y points in a way that maximizes the squared distances between the points and the line(T/F)

False

There is a single definitive test that can be used to establish validity (T/F)

False

To calculate an independent t-test you need three pieces of information from each group. The mean, the standard deviation, and the mode (T/F)

False

Unfortunately, EXCEL cannot perform matched/related/dependent t-test(T/F)

False

When we observe a strong correlation, we know that one variable caused the other(T/F)

False

If we run a chi-square test and find that there is an association between whether someone played baseball(yes or no) and whether someone excels at math (yes or no), we can safely conclude that playing baseball CAUSES one to develop math skills(T/F)

False(one exception to cause-is if you do an experimental trial(if you do two math training programs, then one program may CAUSE better math)-is cause if you don't see random )

A good survey should take about 40 min to complete(T/F)

False-about 20

The intercept in a regression equation represents the point at which the regression line crosses the Y axis when X equals 1(T/F)

False-when it is 0

SD provides ambiguous info when distributions are symmetrical(T/F)

False?

3: In a regression model predictions are called the ___________ variables

Independent

The three forms of reliability are:

Internal consistency, interrater, and stability

The two levels of measurement upon which mathematical operations can be carried out are .......and ......(Fill in the blank)

Interval and Ratio

Could you be at least 99% certain that the store's score of 4.3 was higher than expected?(Y/N)

No

Data such as whether someone is Canadian, Mexican, Peruvian, and German is best coded as Ordinal or Nominal(choose one)

Nominal

To answer the previous question, what part of the 1-2-3 rule can you use?

Part 2

What part of the 1-2-3 rule are you using to test against 95% probability? Part 1 Part 2 Part 3

Part 2

What part of the 123 rules was use to test against 99%

Part 3

One way that Sysco estimated the impact of changes in work attitudes on firm performance was to estimate cost-savings associated with (individual performance or retention numbers)

Retention numbers

The square root of the variance is....

SD

Your company presents data as a z-scores a store has a z-score of a -2.33 on engagements

Significantly Low

(From previous question) If you have a z-value of -.71

Still 26.11%

A contingency table presenting data for all combinations of 2 or more variables is the core of a chi-square test(T/F)

True

A correlation coefficient can take on any value between -1 and 1(T/F)

True

A frequency distribution is a table or graph that indicated how many times a value or score appears(T/F)

True

A key part of running a chi-square test is to contrast expected versus observed frequencies in cells of a contingency table(T/F)

True

A major limitation with the variance is that the variance causes you to lose sight of the original units of measurement(T/F)

True

A normal curve or distribution(not to be confused with a standard distribution) can have any mean and SD(T/F)

True

A normal distribution is symmetrical(T/F)

True

A standardized normal curve always has a mean(T/F)

True

A t-test for a matched/related/dependent sample is based on a different score for each subject(T/F)

True

According to the author, many assumptions about HR programs are based in naive logic(T/R)

True

Another term used to describe a statistical interactions is "moderation"(T/F)

True

Behaviors assessed in surveys should have a clear link to company performance(T/F)

True

Content validity is often established using subject matter experts (T/F)

True

Convergent validity is established by testing whether two or more measures of the same or similar construct are highly related (T/F)

True

Correlations can take on any value between -1 and plus one(T/F)

True

Deviations from the mean will always sum to zero (T/F)

True

EXCEL can estimate a correlation using either the CORREL or PEARSON function(T/F)

True

EXCEL can perform independent t-test(T/F)

True

Generally speaking, one way to increase reliability of a scale is to add more items measuring the scale construct(T/F)

True

If a curve is a normal curve, the total shape is defined by the mean and SD(T/F)

True

If an increase in X is associated with a decrease in Y, the correlation coefficient would be negative(T/F)

True

If you observe a correlation of -0.75, between two variables, you would be correct in stating that the association was strong(T/F)

True

If you really want to make a strong claim that something causes something else, you often have to design an experiment that randomly assigns individuals to different conditions(T/F)

True

If your boss wanted to know if the exempt/non-exempt status of employees differed for male and females, you would be correct to suggest using a chi-square test(T/F)

True

In a standardized normal curve, the values on the x-axis ALSO represent the z-value(T/F)

True

In regression, the residual is the difference between a) the line-based estimate and b) the observed value(T/F)

True

Including negatively worded items is a good idea in survey design(T/F)

True

It is important to consider whether the variances of your two groups are equal when conducting independent t-tests(T/F)

True

It is possible to have a highly reliable score that is still deficient(T/F)

True

On a normal curve, the pt at which the curve begins to turn outward-the pt of inflection-will be one SD away from the mean(T/F)

True

One example of a matched/related/dependent sample is when we measure the same subject at two different times(T/F)

True

One good way to think about a test being significant(Chi-square test, t-test) is to consider whether the pattern you observe differs from chance patterns(T/F)

True

Ordinal data such as Low, Medium, or high is a form of categorical data(T/F)

True

Paper-based surveys appear to produce better data than computer-based surveys(T/F)

True

Reliability indexes range from 0 to 1

True

Reliability indicates the degree to which measurement scores are free of random error(T/F)

True

The coefficient of determination r^2 lets you know how much of the overall variance is explained by the relationship between the x-variable and the y-variable(T/F)

True

The descriptive stats option under the "Data Analysis" add-in is a quick way to get summary statistics such as the mean, median, mode in EXCEL(T/F)

True

The greek term sigma is an abbreviation used to signify that values should be summed(T/F)

True

The mean is more sensitive to extreme values is the median(T/F)

True

The mean, median, and mode all share the same value in the standardized normal curve(T/F)

True

The range is a measure of dispersion(T/F)

True

The statistical symbol Ho represents the null hypothesis(T/F)

True

The value of the slope (b Term) in a regression equation tells how much Y changes with a 1 point change in X(T/F)

True

When there is a fairly even distribution of cases over all the cells of a contingency table there is probably little, if any, association between two variables(T/F)

True

You use N-1 in the denominator of the variance estimate when you have a sample and plan to apply the estimate to a larger population(T/F)

True

The academic literature suggests that cohesion has both a kinship/binding component and a task-focused component. If this is true, are the items above (excluding the item with contamination error) deficient with respect to measuring the construct? (Y/N)

Yes

Current state of the art HR is dominated by state measures(T/F)

false

The EXCEL function to estimate the mean is MEAN

false(its avg)