Reasoning about the design and execution of research

Also statistics!

What is

Positive Control

Negative Control

You’re testing a drug’s ability to kill bacteria:

• Positive control = known antibiotic (you expect bacteria to die).

• Negative control = water (you expect no effect).

• Test = your new drug.

If your positive control doesn't kill the bacteria, your test is invalid — something went wrong.

Because bias is a ____ error in data, only an _____ tool will introduce bias, but an _____ tool will introduce error.

1. Systematic

2. Inaccurate

3. Imprecise

Difference between accuracy and precision? What is their alternative name?

• Accuracy (Validity) = how close a measurement is to the true value.

• Precision (Reliability) = how consistent repeated measurements are.

Draw out a hyperbolic and sigmoidal curve!

Without blinding in an experiment, the ____ would be greatly ____ in the control group but ____ in the treatment group

1. Placebo effect

2. Reduced

3. Present

A cohort study is a form of _____ study

Longitudinal

A relationship posited by an experimental study would be _____ while by an observational study it would be ____.

1. Causal

2. Correlational

What are Hill’s Criteria? Which is the most necessary?

1. Temporality: IV before the DV

2. Strength: More variability in outcome b/c of variability in study variable

3. Dose-dependent: One increase, the other increase

4. Consistency

5. Plausibility/DV Supported by existing info

6. Consideration of alternative explanations: Eliminate other explanations

7. Experiment

8. Specificity: Change is from IV only

9. Coherence: New data agrees with current knowledge

TEMPORALITY

Bias is an error during ___, while confounding is an error during ____.

1. Data collection

2. Analysis (Incorrect relationship is recognized

What is detection bias?

A researcher uses his prior knowledge to actively look for second variable associated with the first variable according to prior research.

Ex: HTN and DM are common in obese people, so a doctor will screen obese patients for DM and HTN more than other patients inflating the proportions

What is the difference between autonomy and respect for persons?

Autonomy: Respect HC decisions on their own behalf

Respect for Persons: Honesty between subject and researcher, informed consent and freedom from coercion mainly in research studies

What is justice in research?

Treat similar patients with similar care and to distribute HC resources fairly

Difference between beneficence and non-maleficence?

Bene- Act in patients best interest

Non-Mal-avoid treatments that causes more harm than good

What are morally relevant differences?

age

Ex: Give liver to child rather than old adult

What is an equipoise?

Knowing one treatment is superior to another but administering them anyway. Goes against beneficence.

If a study chooses participants not part of the target population, what are we violating?

Justice

How can compensatory influence and coercive influence affect a participation?

compensatory influence is one that does not impact the decision to participate, while a coercive influence is one in which the subject loses autonomy to make the decision to participate.

What is the difference between internal validity and external validity?

Internal validity: tendency of the same experiment to produce the same results when repeated, and provides support for causality.

External validity (Generalizability): ability to take the information generated during research and apply it to a larger group

What are considered vulnerable persons?

Children, pregnant women and prisoners

What does an outlier affect?

Mean

What types of data sets are best analyzed using the mean as a measure of central tendency?

The mean is the best measure of central tendency for a data set with a relatively normal distribution. This is why mean doesn’t go well with outliers

Approximately ___ of the distribution is within one standard deviation of the mean, ___ within two, and ___ within three

1. 68%

2. 95%

3. 99%

The interquartile range can be used to determine ____.

Outliers

What is another definition of an outlier?

Any value that lies more than 3 SD from the mean

What’s the difference between a mutually exclusive outcome and a dependent event?

Dependent event is when the outcome of one event effects the other outcome whereas mutually exclusive outcome happens when outcomes cannot happen at the same time (heads+tails)

How do we calculate that at least one of two events will occur?

Do the “or” addition but also subtract the “and” probability

The null hypothesis is a hypothesis that is ___

The alternative hypothesis is __

1. Equal

2. Nondirectional (not equal) or directional (one is greater)

A p value must be between ___ and ____.

0 and 1.

When we know the outcome before a risk factor, the study we would most likely use is a _____ study

Case-control

How is variance related to the correlation coefficient?

square root of variance = R (correlation coefficient)

Do we compare results to a positive control?

NO

What is the range? Where can it be utilized?

The difference between the highest and lowest value data set!

We can use it to find some sort of minimum value for something and we also want to consider if there is any limitations like safety

How do we know a standard deviation is large and therefore make a data point not correct?

1. Relative to the Mean (Coefficient of Variation)

Use the ratio:

CV=Standard Deviation/Mean×100%

• CV < 10% → low variation (small SD)

• CV > 30% → high variation (large SD)