Psych 2220 Final Exam

If it is hypothesized that Canadians will have a lower mean, the researchers may choose to run a 1-tailed test (alpha = 0.05). For a one-tailed test, choose the appropriate null hypothesis.

Canadian adults do not have lower average GNT scores than English adults.

*We calculate a statistical power and find that it is 0.65. This means that if the null hypothesis is ________, we have a ________% chance of rejecting the null hypothesis.

False; 65

*On a test of marital satisfaction, scores could range from 5 to 29. a. What is the full range of data

25

In order to assess the effect of pet ownership on health, a researcher surveys 150 people who own pets and 150 people who do not own pets. The sample being studied is:

300 people

Using the following set of data (which includes 19 values), construct a single stem-and-leaf plot (select the best representation on the Carmen quiz): 3.5 6.5 4.0 3.5 4.5 4.0 3.0 3.5 3.0 3.0 4.0 4.5 3.5 3.0 3.5 3.5 0.5 1.0 3.5

Stem Leaves 0 5 1 0 2 3 0 0 0 0 5 5 5 5 5 5 5 4 0 0 0 5 5 5 6 5

Most psychological research is undertaken with the hope of

rejecting the null hypothesis

Dr. Jameson wanted to know if IQ scores differed between male and female participants in his study. He gave 28 participants an IQ test and then compared IQ scores for gender differences. He hypothesized that there would be a statistically significant gender difference in IQ scores. Contrary to Dr. Jameson's hypothesis, there were no differences in IQ scores between men and women in his study. Which test was Dr. Jameson most likely to use to test his hypothesis?

independent-samples t test

A p level of 0.05 corresponds to a confidence interval of ________%.

95

Assume for a given study that the null hypothesis asserts the expected value of a phenomenon is 100. A research study results in a 95% confidence interval reported as [98.76, 105.24]. What decision would you make based on this confidence interval?

Fail to reject the null hypothesis

Correlation coefficients at or near -1.00 or 1.00 are relatively common.

False

*With which p-level, 0.05 or 0.01, is it easiest to reject the null hypothesis, and why?

It is easier to reject the null hypothesis at the 0.05 level, because it has a larger critical region.

A group of 19 students were surveyed and asked to report: How many hours did you sleep last night? How many classes are you currently taking? a) Is a bar graph the best graph to use for the variables b) Explain your answer to part a (use criteria such as number of variables and scale of measurement to explain your response).

No, because the two variables being measured "hours of sleep" last night and "number of classes" are both scale. Therefore, a scatter plot would have been more appropriate.

A researcher interested in days spent unemployed and feelings of efficacy related to work wonders what impact one's explanatory style might have on the relationship between these variables. She wants to be sure to understand the relation between her variables as clearly as possible. What technique could she use to help address this issue, and what would be accomplished through this method?

Partial correlation would allow her to examine the relationship between length of unemployment and efficacy while removing the influence of the third variable, explanatory style.

If there is an r statistic -5.08, an an alpha level .05, a critical region of (+,-) 1.58, and a p statistic .03. what is your decision?

Reject null, because the r statistic is in the critical region.

A state transportation department is determining the safety of roadways in a locality based on the age of drivers and number of traffic accidents. The department calculates a Pearson correlation coefficient of 0.295. If we determine that the critical value for r is 0.330, what would we conclude?

The age of drivers is not significantly related to the number of traffic accidents.

In a study of the effects of humor on memory, Schmidt (1994) had participants read either humorous sentences or non-humorous sentences and then later tested participants' memory for the sentences. Identify the null hypothesis this study.

There is no difference in memory for humorous and non-humorous sentences.

The proportionate reduction in error is a measure of the:

amount of variance in the dependent variable explained by the independent variable.

The average of all possible split-half correlations is:

coefficient alpha, a measure of reliability.

Geneva believes she is allergic to cats. She has always avoided them and feels ill when she notices one is around. Katarina does not believe Geneva is actually allergic to cats since Geneva seems to have no reaction to the kitten Katarina has secretly been hiding in her room in their shared apartment. Maybe Geneva is remembering the times when she did not feel well and there were cats present and is forgetting the times she has been around cats and felt fine, a phenomenon also known as:

confirmation bias

As part of their work in a research methods class, a group of psychology students devised a survey to assess the relation between stress and health. Each member of the class administered the survey to 10 friends, and the data were then pooled. What method of sampling was used?

convenience sampling

With a p level equal to or above .05 what do we do and how do we notate that

fail to reject p=419, or p>.05

In a single-sample t test, you use the variability of the population to estimate the variability of the sample.

false

The interquartile range is the ________ percent of the data.

middle 50

If the standard error of the estimate is zero, the relation between two variables is:

perfect.

The mean of the population is represented by the symbol ________ and the mean of the sample is represented by the symbol ________.

μ; M

The symbol for the standard error is:

σM.

You are playing with a standard deck of 52 playing cards. Each time you draw one card from the deck and then you put the card back and reshuffle the deck before choosing another card. Drawing a card from the deck is a(n) ________.

Trial

A z test result of 4.12 would allow us to reject the null hypothesis, assuming we were conducting a two-tailed test with critical values of +1.96 and -1.96.

True

On a histogram the y-axis depicts ________, but on a bar graph the y-axis depicts ________.

counts or frequencies; the dependent variable

A health psychologist studies food intake by recording the type of food consumed by participants at lunch (such as: meat, carbohydrate, vegetable, dairy). The scale of measurement for type of food would be:

nominal

If all the points on a scatterplot fall on a single line:

the relation between the variables is perfect.

The phrase statistically significant means that:

the research result was unlikely to have occurred by chance.

A researcher studies 45 volunteer citizens from a small community and asks them about the amount of caffeine (in milligrams) they ingest before and after lunch each day. Two measures are taken from each participant for a total of 90 data points. The degrees of freedom for this paired-samples study are:

44

b. What would the interval size be if we wanted five intervals and the scores range from 5-29?

5

Suppose we compute the standard deviation and find that it is equal to 5.50. This number means that the numbers in the sample deviate, on the average:

5.50 units from the mean.

Researchers who study the process of reading in children have discovered that better readers make smoother eye movements across the page (i.e., there is a positive correlation between reading skill and the smoothness of the eye movements). Which of the following statements is a possible cause of the correlation between smooth eye movements and reading ability?

All of these options are possible causes of the correlation.

The population mean for a standardized final exam in a course is an 83%. After taking a large sample from the new freshmen class, a 99% confidence interval of [84.81, 86.32] is calculated. If a 95% confidence interval were calculated based on the same data, it would _____ compared to the interval listed above.

Be shorter (include fewer values)

Why does random assignment help control for confounding variables?

By randomly assigning people to groups, individual differences that may influence the dependent variable are randomly distributed throughout the conditions, rather than being systematically related to the independent variable.

*A researcher investigates if people's motivation at work changes following a vacation. He follows 8 individuals who went on week-long tropical vacations, assessing their work-related motivation before and after the vacation. Given the following confidence interval [-0.05, 2.76], make a decision about the hypotheses.

Fail to reject the null hypothesis because zero falls in the confidence interval.

If we know that the percentage of scores falling between the mean and a z score of 0.40 is 15.54, then the percentage of scores falling between the mean and a z score of -0.40 is 34.46.

False

Based on the equation: Yhat = 2.306 + 3.7X1 -2.9X2, how would you interpret the slope for X2?

For every one unit increase in the independent variable X2 the Yhat dependent variable will decrease by 2.9 while controlling for all other variables

A researcher is worried that performance on a fitness test will improve simply because of repeat exposure to the test, which will cloud his ability to assess the impact of two unique fitness programs he wants to study (Program 1 and Program 2). Which of the following design options includes counterbalancing?

He could vary the order of the educational interventions such that half of the participants experience Program 1 first and the other half experience Program 2 first.

*select the symbolic version of the null hypothesis: A student wonders if grades in a class are in any way related to where a student sits in the classroom. In particular, do students who sit in the front row get better grades, on average, than the general population of students?

Ho: μ(front_row_students) ≤ μ(general_pop_students)

select the symbolic version of the null hypothesis: Cell phones are everywhere, and we are now available by phone almost all of the time. Does this translate into a change of the closeness of our long-distance relationships?

Ho: μ(with_cellphones) = μ(without_cellphones)

A power analysis reveals that the study being run has low power. Which of the following methods is not an appropriate way to increase statistical power?

Increase the variance of the distributions.

*When creating a frequency table for the example below, select whether you should use individual data values or grouped data. Number of siblings for 45 college students

Individual data values

The Lee family is looking to buy a house in one of two suburban areas just outside a major city, and the air quality is a top priority for them. One suburb advertises use of hybrid cars and solar panels, while the other area focuses on its convenient bus routes and availability of Hummer dealerships. Is the mean or median the better measure to use for deciding which area has better air quality? (Hint: both these populations are skewed.)

Median; because it is less biased by skewness being dependent on the middle score

It becomes progressively easier to declare statistical significance as we increase sample size.

Raising sample size reduces standard error of the mean (in z-tests) and increases statistical power, making it easier to reject the null hypothesis. A test is significant when we reject the null.

*Researchers are interested in depressed individuals who are not responding to treatment. For their study, the researchers sample 18 patients from their own private clinics whose depression had not responded to treatment. Half received one intravenous dose of ketamine, a hypothesized quick fix for depression; half received one intravenous dose of placebo. Far more of the patients who received ketamine improved, as measured by the Hamilton Depression Rating Scale, usually in less than 2 hours, than patients on placebo. Would random assignment be possible to use?

Random assignment would be possible because we can randomly assign a number to each participating patient signifying which of the two groups they belong to and give them the corresponding medication.

Suppose that for a particular study the expected value of a phenomenon is 0.00. The study results in a reported 99% confidence interval of [0.05, 0.45]. What decision do you make based on the 99% confidence interval?

Reject the null because the population mean is not within the confidence interval.

The population mean for a college Introduction to Biology quiz is 20 points. A sample of 35 high school seniors earns an average of 16.53 points. A 99% t confidence interval for the true mean of high school seniors is [13.85, 19.21]. Based on this confidence interval, what would your decision be in regard to a 2-tailed null with an alpha level of 0.01?

Reject the null, because 20 is not in this interval.

Dr. Baker designed an experimental study to assess potential differences between science students and art students on a math reasoning abilities test. Dr. Baker found a mean difference in math performance between science and art students. On average, art students performed higher on the math reasoning test compared to the science students. Dr. Baker's findings support which hypothesis?

Research Hypothesis

*As sample size increases, the test statistic increases because:

Standard error decreases

A new GRE prep class finds a sample from a recent session of their class has average scores that produce an effect size of d = -0.25 compared to the known GRE average. Interpret this effect size

The effect size is minimal (small) and below the mean for the two groups. Cohen's d is the is the difference between the two groups mean's divided by the standard deviations.

*A researcher conducts a single-sample t test and finds statistical significance at the 0.01 level. The effect size is then calculated and found to be 0.50. What might you conclude about the findings?

The results are both statistically and practically significant in this case, as the effect size indicates a medium effect, and the 0.01 level of significance is rather impressive*.

Twenty college students experience the effects of alcohol on reaction time. They perform very basic timed responses in a driving simulator both before and after consuming several alcoholic beverages. The researcher collects a reaction-time result for each of the 20 students before and after intoxication, for a total of 40 measures. What is the 2-tailed null hypothesis for this paired-samples study?

There is no difference in reaction time before and after consuming alcohol

A paired-samples t test is reported as t(20) = 3.58, p < 0.05, d = 1.89. Interpret the effect size for this statistical test.

This is a large effect, indicating the difference likely has practical importance.

A researcher is interested in how students at OSU feel about smoking on campus. The researcher goes out on the Oval and collects responses of students that walk by. Suppose that some time later new policies were put in place and the researcher asked students about their attitudes before and after the implementation of these policies. The researcher makes the hypothesis that "Students' positive attitudes towards smoking will decline due to the new policy." The study results indicate that positive attitudes towards smoking do not significantly decrease as a result of the new policies. However, in reality, positive attitudes towards smoking did decrease. This is a _________ error.

Type II

Garret thinks his car battery is becoming weak. He doesn't want to be stranded on the highway, so he tests his battery and it tests fine. But when Garret begins his trip on the highway, the battery dies 20 miles down the road. The test results must have been wrong! What type of error is this?

Type II

By increasing statistical power, the probability of making a ________ error is ________.

Type II; decreased

Why are effect sizes rather than test statistics used when comparing study results?

Using effect sizes, which are not affected by sample size, rather than test statistics, which are influenced by sample size, ensures a fair comparison.

Under what conditions is it permissible to proceed with a hypothesis test even though the assumption that the population distribution is approximately normal is violated?

We have a sample size greater than 30

Assume that we find a positive correlation between the number of hours students spend studying for an exam and their grade on the exam. If we calculate the regression equation for these data and find that the y intercept is 65, what conclusion can we draw?

When students do not study at all, we would predict a score of 65 on the exam.

Researchers provided students with a survey that asked if they typically read their textbook before material was covered in class (possible answers: yes or no) and to report their most recent exam grade (ranging from 0-100%). The most appropriate graph to represent this data would be a _____, because _____.

bar graph; one variable is nominal and one variable is scale

A student poll on your university Web site for athletics asks, "How high do you think the basketball team will finish this year: 1st, 2nd 3rd, 4th, or 5th?" An article posted a week later reports "Students predict basketball team to finish within the top 5 out of 16 teams in the conference." What kind of manipulation occurred here?

biased scale lie

state whether the trials are independent or dependent. 9-year-old Lucy gets to pick from the candy bag, filled with colored jelly beans, every time she helps around the house. She has been picking (and eating!) candy for 5 days in a row and each time she has picked a different color jelly bean. The candy bag was only filled one time and no more candy has been added since then

dependent

*Wendy is a Weight Watchers group leader. To get a better idea of how to help those she will be working with to achieve their weight-loss goals, she wishes to know the average weight-loss goal of the individuals in her group. What kind of statistic should Wendy use?

descriptive

The comparison distribution for an independent-samples t test is a distribution of:

differences between means.

The United States Department of Homeland Security Threat Advisory System measures threat as severe, high, elevated, guarded, or low. In this system, threat is a ________ variable.

discrete and ordinal

Stem-and-leaf plots offer an advantage over histograms and frequency polygons in that they allow us to:

easily compare two groups of data together

For a simple linear regression, the standardized regression coefficient (Beta) is:

equal to the Pearson correlation coefficient.

Nineteen ninety-five data from over 1,000 colleges published in U.S. News & World Report were used to predict the percent of alumni who donate to a college from the average SAT score of students attending the college. The resulting regression equation was Ŷ = -29.29 + 0.05(X). This regression indicates that:

for every one-point increase in SAT scores, a college can expect 0.05% more of its alumni to donate.

A researcher has conducted a study on logical reasoning using a sample of college freshmen and is concerned that the results of the study may not accurately depict the logical reasoning of college seniors. The researcher's concern can be described as a concern about the ________ of the study results.

generalizability

A z score of -1.0 is ________ compared to a z score of -2.0.

higher

The practical use of statistical power is that it informs you the researcher:

how many participants are needed to conduct a study that will produce quality data that you can trust.

To test whether a chemical found in many common plastic items is toxic, researchers fed different amounts of the chemical (from 0 to 20 micrograms) to mice and later performed a blood test to determine the amount of toxins present in each mouse. To create a scatterplot of the data, which variable should be listed on the x-axis?

independent variable, which is the amount of the chemical given to the mice

At a sample size of infinity, the t distribution:

is identical to the z distribution

Imagine that a study of memory and aging finds that younger participants correctly recall 55% of studied words, older participants correctly recall 42% of studied words, and the size of this effect is Cohen's d = 1.5. According to Cohen's conventions for interpreting d, this effect is:

large

When running a single-sample t-test, you should fail to reject the null hypothesis when your test statistic (t) is:

less extreme than your critical value(s).

Based on research with her patients, Dr. Sabine knows that the correlation coefficient between scores on an anxiety scale and comfort at a social gathering is 0.35. According to guidelines established by Cohen, how could we characterize the strength of this relationship?

medium

Assume a researcher wants to understand usage pattern of buses in Columbus. She passes out a survey on the bus where one question is "Which bus route do you use most often?". The following subset of a larger data set is X = {10, 11, 2, 8, 21, 2, 2, 8, 18, 14, 10}. (NOTE: each number corresponds to a bus route) Pick all measures of central tendency that are appropriate for this data

mode

After everyone finished the race, the organizers noted that while most runners finished in 6-8 minutes, a few participants actually walked and finished in 15 minutes. This distribution could considered _____.

positively skewed

For a set of 50 numbers a researcher calculates the following measures of central tendency: mode = 5, median = 40, mean = 60. This distribution of this data would be considered:

positively skewed

A researcher was interested in assessing whether a new medication had negative side affects on reaction time. He performed an experiment on a group of rats. One group of rats received the medication, and the other group of rats did not. He then measured reaction time differences between the two groups on a series of tasks. In this experiment, the dependent variable is:

reaction time

Researchers are interested in depressed individuals who are not responding to treatment. For their study, the researchers sample 18 patients from their own private clinics whose depression had not responded to treatment. Half received one intravenous dose of ketamine, a hypothesized quick fix for depression; half received one intravenous dose of placebo. Far more of the patients who received ketamine improved, as measured by the Hamilton Depression Rating Scale, usually in less than 2 hours, than patients on placebo. What decision did the researchers likely make?

reject the null

What is the research hypothesis when testing for significance using the Pearson correlation coefficient?

rho ≠ 0

To efficiently conduct an alumni survey, a university collects data on all those who attend the annual alumni reunion on campus. What type of manipulation is this?

sneaky sample lie

When statisticians describe pooling the variances, they mean:

taking the average of the two variances accounting for sample size.

*According to ________, as sample size increases, the distribution of ________ assume a normal shape.

the central limit theorem; sampling scrores

A z-score is:

the number of standard deviations a particular score is from the mean.

The statement "It is hypothesized that men and women will differ on reaction time measures" best illustrates a:

two-tailed test.

People debate whether standardized tests, such as the ACT and SAT, are predictors of college performance. This is essentially a debate about:

validity

Mark is looking at a histogram depicting years of education. In looking at the histogram, the x-axis most likely represents the ________ while the y-axis most likely represents the ________.

values of the variable "years of education"; frequencies or number of people

In a reanalysis of published studies, Twenge and Im (2007) found that for the time period 1958-2001, the need for social approval of people in the United States was negatively correlated with the U.S. divorce rate during the same period (the correlation coefficient was -0.38). This correlation means that:

when the need for social approval was high, divorce rates were low.

Some researchers reported a trend for women to lose an unhealthy amount of weight between the time of their wedding dress purchase (represented by "p") and their actual wedding day (represented by "w"). The question of interest is: Do women lose weight for their wedding day? Select the appropriate 1-tailed null hypothesis:

μw ≥ μp