Biostats 1- Basic Concepts Flashcards by Daniel Guck

This is the variation in the values of a sample statistic when multiple samples are taken from the same population.

Random sampling variation

How well did you know this?

Not at all

Perfectly

This is the selection of subjects from a population to participate in a study.

Sampling

How well did you know this?

Not at all

Perfectly

What is the goal of sampling?

To obtain a representative sample that mimics characteristics of the population.

How well did you know this?

Not at all

Perfectly

This is the type of sampling where each individual in the population has an equal chance of being selected to participate.

Random sampling

How well did you know this?

Not at all

Perfectly

This is the table you use in random sampling where u assign each person a # from 100 and then to randomly select from it.

Random # table

How well did you know this?

Not at all

Perfectly

This is the type of sampling where u select a starting point and then select every K element in the population.

Systematic sampling

How well did you know this?

Not at all

Perfectly

To do a systematic sampling, how do arrange individuals?

In some ordering scheme and then selecting at regular intervals through that ordered list.

like interviewing every 5th pt for 1 week in a study.

How well did you know this?

Not at all

Perfectly

This is the type of sampling where u subdivide the population into subgroups that share the same characteristic, and then draw a random sample from each stratum.

Stratified sampling

How well did you know this?

Not at all

Perfectly

This is the type of sampling where u use results that are readily available or easily accessible, like phone calls and mall questionairres.

Convenience sampling

How well did you know this?

Not at all

Perfectly

This is the type of sampling where u divide the population into sections and then randomly select some of those clusters (choose all members of the selected cluster)

Cluster sampling

How well did you know this?

Not at all

Perfectly

So what’s the difference between stratified and cluster sampling? both take sections of a popualtion…

in cluster, you sample ALL of the people in the selected strata

in stratified, you select a few from each strata.

How well did you know this?

Not at all

Perfectly

What are the 2 ways u can classify data into variables?

Sacle of measurement

Arrangement in a study design

How well did you know this?

Not at all

Perfectly

What are the 2 ways to split up the scale of measurement when classifying data?

Qualitative and quantitative

How well did you know this?

Not at all

Perfectly

Further down the rabbit hole… What are the 2 ways to split up the qualitative (categorical) variables?

Nominal and ordinal.

How well did you know this?

Not at all

Perfectly

Which categorical (qualitative) variable does this belong under (nominal or ordinal)?

Republican, Democrat, Green party

Nominal

theres no order to the political parties

How well did you know this?

Not at all

Perfectly

Which categorical (qualitative) variable does this belong under (nominal or ordinal)?

Hotness of a chili pepper, age group, pain rating

Ordinal

theres an order, so u rank the values in order.

How well did you know this?

Not at all

Perfectly

If there is a categorical (qualitative) version, what is the other type of scale of measurement when classifying data?

Continuous (quantitative) variables

How well did you know this?

Not at all

Perfectly

What are the 2 subsections of quantitative variables?

Intervals and Ratios

How well did you know this?

Not at all

Perfectly

Which continuous (quanitative) variable does this belong under (Interval or Ratio)?

Temperature

Interval

can be negative

How well did you know this?

Not at all

Perfectly

Which continuous (quanitative) variable does this belong under (Interval or Ratio)?

BPM, BP, height, weight

Ratio

never negative

How well did you know this?

Not at all

Perfectly

Which values are the input and manipulated in a study design (independent or dependent variables)?

Independent variables

How well did you know this?

Not at all

Perfectly

Which axis are independent variables always placed on a graph (X or Y)?

X-axis

How well did you know this?

Not at all

Perfectly

This is the extento to which an instrument will provide the same values when repeatedly re-measuring.

Reliability

How well did you know this?

Not at all

Perfectly

This is the subtype of reliability where there is different people, same test.

Inter-rater reliability

between people, like people checking their weight

How well did you know this?

Not at all

Perfectly

This is the subtype of reliability where there is the same people, same test, different times.

Test-retest | like checking ur own weight every morning on the same scale

This is the subtype of reliability where there are different questions, same constuct

Internal consistency

This is the extento to which a measurement, tool, or study is an accurate representation of the concept it is intended to measure.

Validity

Can validity be quantified?

This is the subtype of validity where the extent to which the actual measurements made by investigators are truly representative of what they are observing and measuring

Internal validity

So to whom does external validity apply?

Populations, settings, or treatments.

This is the type of bias where there is systematic deviation form reality, and the samples or relsults are NOT representative of the population.

Statistical bias

This is the type of statistical bias which occurs when the sample isnt representative of the population (not equal change of being selected), which lowers external validity.

Selection bias | Dewey defeats Truman

What are the 2 ways of having selection bias?

``` Migration bias (drop out) non-response bias (dont participate) ```

THis is the type of statistical bias which lacks proper blinding (investigator bias).

Review bias | like the MMR/autism study

What is the most common way of having review bias?

Where only 1 person collects data/test results | reader/verification bias

This is the type of statistical bias where hte pt or subject skews results due to knowledge of being tested.

Response/artificial bials

What are the 3 preventative measure against bias?

Randomization Double-blind experiments Proper training in research

This is a variable or factor that is anything other than the treatment that can cause or preclude the outcome of interest.

Confounding variable/factor | like in stress vs BP, know that other things that can change BP, so u wana reduce confounding factors

This is the deviation of a measured value from the true meaning.

Measurement error.

Which type of measurement error is harder to correct, random or systematic errors?

Random errors cuz u dont know whats causing the problem

This is the discrepancy between the observed result and the true value. ``` A. Accuracy B. Bias C. Confounding D. Measurement error E. Precision F. Reliability G. Validity ```

D. measurement error

This is the lack of error, getting the correct result. ``` A. Accuracy B. Bias C. Confounding D. Measurement error E. Precision F. Reliability G. Validity ```

A. accuracy

This is the absence of a random error. ``` A. Accuracy B. Bias C. Confounding D. Measurement error E. Precision F. Reliability G. Validity ```

E. precision

This is a systematic error- independent of study size. ``` A. Accuracy B. Bias C. Confounding D. Measurement error E. Precision F. Reliability G. Validity ```

B. bias

This is the absence of bias (or absence of all error) ``` A. Accuracy B. Bias C. Confounding D. Measurement error E. Precision F. Reliability G. Validity ```

G. validity

These influence the outcome of interest. ``` A. Accuracy B. Bias C. Confounding D. Measurement error E. Precision F. Reliability G. Validity` ```

C. Confounding

This is a table listing all the possible values and the # of occurences of each value in the sample.

Frequency distribution

What kind of chart can u use to visualize the freuqency distribution for categorical variables?

Bar chart

How many slices should a pie chart not exceed?

5-6

This is the graph that can depict frequency distributions for quantitative data.

Histograms

What plot can visualize the replationship between 2 quantitative variables?

Scatter plot

Should you use stem and leaf plots for large or small data sets?

Small data sets

In a box plot, what is depicted by the ends of the box?

interquartile range

In a box plot, what is depicted by the midline inside the box?

Median

In a box plot, what is depicted by the + symbol in the box?

Mean

In a box plot, what is depicted by the whiskers?

2 SD from the mean

In a box plot, what is depicted by the dots beyond the whiskers?

Outliers

This is the measure where u add up all the data and divide by the total.

Mean

This is the measure where u find the 1/2 way pt, where there is an equal # of variables on both sides.

Median

This is the measure of the most frequent data in a set.

Mode

This is the measure of how much variation exists in the data, and is comprised of the range, interquartile range, and the variance.

Measure of Dispersion

What is the eqn for the variance?

E(obs - mean)^2/n

If you take the square root of the variance, what do u get?

standard deviation

This is the amt of variation in means of samples when repeated sampling is used from a popualtion.

Standard error

What % of data is included within 1 SD in a bell curve?

68%

What % of data is included within 2 SDs in a bell curve?

95%

What % of data is included within 3 SDs in a bell curve?

99.7%

If a distribution curve has a long right tail, is it a positive or negative skew?

Positive

Which do u report in a skew: mean, median, or mode?

Median

This is the measure of peekedness of a curve.

Kurtosis

What is it called when a curve is peeked > 0?

Leptokurtic

What is it called when a curve has 0 kurtosis?

Mesokurtic

What is it called when a curve is peeked < 0?

Platykurtic like a platypus. cuz it's flat.

What are the 5 abuses of descriptive statistics?

``` Bad samples Small samples loaded questions ("does this dress make me look fat?") Misleading graphs Precise numbers ```