BILD 5 PRACTICE TESTS

What is the relationship between cognitive biases and the scientific
method?

all humans have biases, to minimize bias, we must follow the scientific method

You have the following set of values: c(12, 15, 8, 5, 15, 16, 9, 10, 97).
Which single descriptive statistic would most accurately describe the center of the entire set of
values? Explain.

the median- its robust (resistant to change) from the outlier

what is the difference in standard error and standard deviation

standard error is the variation of sample means while standard deviation is the variation of the data values

what is the relationship between the 95% Confidence interval and the standard deviation

A 95% CI is 2 standard deviations away from the mean

You flip a coin a bunch of times and use a Chi-squared goodness of fit
test to see if it’s a fair coin. You get a p-value of 0.12. In your own words, what does that p-value
mean specifically?

if the coin is fair, there is a 12% chance that you would get a result at least as extreme

Why would a responsible researcher collect a small amount of data when
beginning a project knowing that they will not use that data in the final analysis?

to conduct a pilot study

The typical alpha is 0.05. Name one statistical and one ethical situation
in which the standard alpha is too high

ethical- stakes are too high

statistical- multiple tests

Your friend tells you that a new drug screening test for Olympic athletes
is 99% accurate when athlete has been doping and 98% accurate when the athlete has not been
doping. Thankfully, you have data that athletes rarely dope (1 in 200). Your friend is so excited
since he says it will be easy to catch all dopers with a single test. Using Bayesian thinking, what
would you say to your friend?

Even if tests are accurate, a positive test includes true positives and false positives. If an event is sufficiently rare then the false positives will make up a large portion of positive test results

In the last two lectures we talked about some of the problems with modern science
culture and potential solutions. Name one of those problems and a specific way you can help
combat it over the next ten years.

• AI usage

What is one similarity and difference between a linear regression and a logistic regression?

similar- they both are predictive modeling techniques

difference- linear regression predicts a data point while logistic predicts a probability

Which scenario is most appropriate for a repeated-measures design?
A. Measuring plant growth under different light conditions across time.
B. Surveying 100 random individuals about their dietary habits.
C. Comparing exam scores between two separate sections of the same course.
D. Collecting soil samples from random plots in a large field experiment.

a

Which best defines ‘blocking’ in experimental design?
A. Randomly assigning subjects to different treatments.
B. Grouping subjects by a known variable (e.g., age) to control its effects.
C. Splitting an experiment into multiple time periods.
D. Combining multiple factors in a single experiment to see interactions.
E. Blocking out certain demographics, so they are never sampled from the population

b-Grouping subjects by a known variable (e.g., age) to control its effects.

Which of the following is not one of the usual assumptions of parametric tests like ANOVA and
t-tests?
A. Samples are drawn from populations with equal variances.
B. The residuals are normally distributed.
C. The sample size must be identical across all groups.
D. Observations are independent of each other.

c-The sample size must be identical across all groups.

Which of the following is NOT a common reason to conduct a pilot study?
A. A full-scale experiment repeated multiple times.
B. A smaller, preliminary experiment to test methods and feasibility.
C. A mandatory government-mandated trial before a larger study.
D. A study focusing on aviation biology

a- A full-scale experiment repeated multiple times.

Which of these statements about statistical power is most accurate?
A. It is determined solely by the sample size.
B. It represents the probability of rejecting the null hypothesis when it is truly false.
C. Power can never exceed 50% in biological experiments.
D. Power decreases as effect size increases.
E. It is equal to probability of a false negative.

b - It represents the probability of rejecting the null hypothesis when it is truly false.

In a field study comparing two fertilizers, each applied to separate groups of plants, which
scenario poses a risk of pseudoreplication?
A. Each plant is measured once in each treatment group.
B. Multiple independent fields are randomly assigned one fertilizer or the other.
C. Each row within one field is labeled a replicate, but the entire field receives a single fertilizer.
D. An equal number of plants is randomly selected from each treatment group for
measurement.

c-Each row within one field is labeled a replicate, but the entire field receives a single fertilizer.

You are planning a study involving students who will take a five-minute survey asking
demographic and health related questions. Which of the following is most critical to address
first, before beginning the experiment?
A. Performing preliminary statistical analyses on pilot data.
B. Having participants sign an informed consent and securing IRB approval.
C. Optimizing sample size with a power analysis.
D. Scheduling follow-up appointments to ensure no subject drops out.

b-Having participants sign an informed consent and securing IRB approval.

A biologist wants to assess whether fertilizer type affects yield in wheat fields known to vary in
soil nitrogen levels. Which sampling design best prevents confounding effects of soil nitrogen
on fertilizer outcomes?
A. Randomly assigning fertilizer treatments to entire fields without recording nitrogen
differences.
B. Stratifying fields by soil nitrogen level and then randomly assigning each stratum to both
fertilizers.
C. Applying each fertilizer to fields chosen by convenience near the lab.
D. Mixing both fertilizers in each field for uniform coverage

b-Stratifying fields by soil nitrogen level and then randomly assigning each stratum to both
fertilizers.

A researcher measures the same individuals’ insulin levels before and after a new dietary
intervention. Which condition must be checked prior to applying a paired t-test?
A. Both samples must come from different individuals.
B. The difference in insulin measurements (pre vs. post) should be approximately normally
distributed.
C. The standard deviations of the two groups must be identical.
D. At least 200 individuals are needed for a valid paired t-test.

b-The difference in insulin measurements (pre vs. post) should be approximately normally
distributed.

A botanist collects the mean height of 50 randomly selected plants. It appears that the plant
height has a uniform distribution between 1-2m. According to the central limit theorem, which
statement best describes the distribution of sample means if she samples the data 1000 times?
A. The sample means will cluster near the population mean and approximate a normal
distribution, regardless of each plant’s individual distribution.
B. Each sample mean will match the population mean exactly if the sample size is at least 30.
C. The distribution of sample means will be uniform between the smallest and largest plant in
the population.
D. The sample means will only be normally distributed if the original plant heights are
themselves perfectly normal.

A- The sample means will cluster near the population mean and approximate a normal
distribution, regardless of each plant’s individual distribution.