Evidence Based Medicine Flashcards Preview

BMS > Evidence Based Medicine > Flashcards

Flashcards in Evidence Based Medicine Deck (171)
Loading flashcards...
1
Q

Allows us to draw from the sample, conclusions about the general population

A

Statistics

2
Q

An efficient way to draw conclusions when the cost of gathering all of the data is impractical

A

Taking Samples

3
Q

Assume that an infinitely large population of values exists and that your sample was randomly selected from a large subset of that population. Now use the rules of probability to

A

Make inferences about the general population

4
Q

States that the sampling distribution of the mean of any independent, random variable will be normal or nearly normal, if the sample size is large enough

A

The Central limit theorem

5
Q

What does the Central Limit Theorem say?

A

The sampling distribution of the mean of any independent, random variable will be normal or nearly normal, if the sample size is large enough

6
Q

If samples are large enough, the sample distribution will be

A

Bell shaped (Gaussian)

7
Q

Statistics come in what two basic flavors?

A

Parametric and Non-parametric

8
Q

A class of statistical procedures that rely on assumptions about the shape of the distribution (i.e. normal distribution) in the underlying population and about the form or parameters (i.e. mean and std. dev) of the assumed distribution

A

Parametric Statistics

9
Q

A class of statistical procedures that does not rely on assumptions about the shape or form of the probability distribution from which the data were drawn

A

Non-parametric Statistics

10
Q

Summarize the main features of the data without testing hypotheses or making any predictions

A

Descriptive statistics

11
Q

Descriptive statistics can be divided into what two classes?

A

Measures of location and measures of dispersion

12
Q

A typical or central value that best describes the data

A

Measures of location

13
Q

What are the measures of location?

A
  1. ) Mean
  2. ) Median
  3. ) Mode
14
Q

Describe spread (variation) of the data around that central value

A

Measures of dispersion

15
Q

What are the measures of dispersion?

A
  1. ) Range
  2. ) Variance
  3. ) Std. Dev
  4. ) Std. Error
  5. ) Confidence Interval
16
Q

No single parameter can fully describe the distribution of data in the

A

Sample

17
Q

The sum of the data points divided by the number of data points

  • More commonly referred to as “the average”
  • Data must show a normal distribution
A

Mean

18
Q

What are often better measures of location if the data is not normally distributed?

A

Median and Mode

19
Q

The value which has half the data smaller than that point and half the data larger

A

Median

20
Q

When choosing the median for odd number of data points, you first

A

Rank the order, then pick the middle #

21
Q

When choosing the median for even number of data points, you

A
  1. ) Rank the numbers
  2. ) Find the middle two numbers
  3. ) Add the two middle numbers and divide by 2
22
Q

Less sensitive for extreme data points and is thus useful for skewed data

A

Median

23
Q

The value of the sample which occurs most frequently

A

Mode

24
Q

The mode is a good measure of

A

Central Tendency

25
Q

Not all data sets have a single mode, some data sets can be

A

bi-modal

26
Q

On a box plot, 50% of the data falls between Q1 (25th percentile) and Q3 (75th percentile), the area encompassing this 50% is called the

A

Interquartile range (= Q3-Q1)

27
Q

Used to display summary statistics

A

Box plots

28
Q

To find the quartiles, put the list of numbers in order, then cut the list into four equal parts, the quartiles are at the

A

Cuts

29
Q

The second quartile is equal to the

A

Median

30
Q

Do not provide information on the spread or variability of the data

A

Measures of location

31
Q

Describe the spread or variability within the data

A

Measures of dispersion

32
Q

Two distinct samples can have the same mean but completely different levels of

A

Variability

33
Q

The difference between the largest and the smallest sample values

-Depends only on extreme values and provides no information about how the remaining data is distributed

A

Range

34
Q

Is the range a reliable measure of the dispersion of the whole data set?

A

No

35
Q

The average of the square distance of each value from the mean

A

Variance

36
Q

Makes the bigger differences stand out, and makes all of the numbers positive, eliminating the negatives, which will reduce the variance

A

Squaring the Variance

37
Q

When calculating the variance, what is the difference between using N vs. N-1 as the denominator?

A

N gives a biased estimate of variance, where as (N-1) gives an unbiased estimate

38
Q

In the calculation for variance, what does N represent?

A

N = size of population (biased)

39
Q

In the calculation for variance, what does (N-1) represent?

A

(N-1) = size of the sample (unbiased)

40
Q

The most common and useful measure of dispersion

A

Standard deviation

41
Q

Tells us how tightly each sample is clustered around the mean

A

Standard deviation

42
Q

When samples are tightly bunched together, the Gaussian curve is narrow and the standard deviation is

A

Small

43
Q

When the samples are spread apart, the Gaussian curve is flat and the standard deviation is

A

Large

44
Q

Means and standard deviations should ONLY be used when data are

A

Normally distributed

45
Q

How can we determine if the data are normally distributed?

A

Calculate the mean plus or minus twice the standard deviation. If either value is outside of the possible rage, than the data is unlikely to be normally distributed

46
Q

Approximately what percentage of data lies within:

  1. ) 1 standard deviation of the mean
  2. ) 2 Standard deviations of the mean
  3. ) 3 Standard deviations of the mean
A
  1. ) 68.3%
  2. ) 95.4%
  3. ) 99.7%
47
Q

If data is skewed, we should use

A

Median

48
Q

What are two more sophisticated, yet more complex, methods of determining normality?

A

D’Agostino & Pearson omnibus and Shapiro-Wilk Normality tests

49
Q

D’Agostino & Pearson omnibus and Shapiro-Wilk Normality tests are not very

A

Useful

50
Q

What we want is a test that tells us whether the deviations from the Gaussian ideal are severe enough to invalidate statistical methods that assume a

-Normality tests don’t do this

A

Gaussian distribution

51
Q

How can we determine whether our mean is precise?

A

Find the Standard Error

52
Q

A measure of how far the sample mean is away from the population mean

A

Standard error

53
Q

The standard error of the mean (SEM) gets smaller as

A

Sample size gets larger

54
Q

If the scatter in data is caused by biological variability and you want to show that variability, use

A

Standard Deviation (SD)

55
Q

If the variability is caused by experimental imprecision and you want to show the precision of the calculated mean, use

A

Standard Error of the mean (SEM)

56
Q

Say we aliquot 10 plates each with a different cell line and measure the integrin expression of each, would we want to use SD or SEM?

A

SD

57
Q

Say we aliquot 10 plates of the same cell line and measure the integrin expresion of each, would we want to use SD or SEM?

A

SEM

58
Q

An estimate of the range that is likely to contain the true population mean

-combine the scatter in any given population with the size of that population

A

Confidence intervals

59
Q

Generates an interval in which the probability that the sample mean reflects the population mean is high

A

Confidence intervals

60
Q

Means that there is a 95% chance that the confidence interval you calculated contains the true population mean

A

95% confidence interval

61
Q

If zero is included in a confidence interval for a change in a disease due to a drug, then it means we can not exclude the possibility that

A

There was no true change

62
Q

An observation that is numerically distant from the rest of the data

A

An outlier

63
Q

Can be caused by systematic error, flaw in the theory that generated the data point, or by natural variability

A

An outlier

64
Q

What is one popular method to test for an outlier?

A

The Grubbs test

65
Q

How do we use the Z value obtained by the Grubbs test to test for an outlier?

A

Compare the Grubbs test Z with a table listing the critical value of Z at the 95% probability level. If the Grubbs Z is greater than the value from the table, then you can delete the outlier

66
Q

To test for an outlier, we compare the Grubbs test Z with a table listing the critical value of Z at the 95% probability level. If the Grubbs Z is greater than the value from the table, then the P value is

A

Less than 5% and we can delete the outlier

67
Q

What constitutes “good quality” data

A

Data must be: reliable and valid

68
Q

What measurements assess data reliability?

A

Precision, accuracy, repeatability, and reproducibility

69
Q

In order for the data to be valid, it must be

A

Compared to a “gold standard,” generalisable, and credible

70
Q

The degree to which repeated measurements under unchanged conditions show the same results

A

Precision

71
Q

High precision results in lower

A

SD

72
Q

The degree of closeness of measurements of a quantity to that quantity’s true value

A

Accuracy

73
Q

High accuracy reflects the true

A

Population mean

74
Q

Repeatability is the same as

A

Precision

75
Q

The ability of an entire experiment or study to be duplicated either by the same researcher or by someone else working independently

-The cornerstone of research

A

Reproducibility

76
Q

The extent to which a concept, conclusion, or measurement is well-founded and corresponds accurately to the real world

A

Validity

77
Q

Assuming that data collected on small samples are indicative of the population, sampling errors (bias, size, etc), and instrument errors are all threats to

A

Validity

78
Q

The generalizability of a study is called it’s

A

External validity

79
Q

Thalidomide was tested on rodents and showed no effects on limb malformations. However, the effects on humans were very pronounced. This is an error in

A

External validity

80
Q

Many studies using single cell lines are no longer

A

Acceptable

81
Q

Are the methodologies acceptable? Do the investigators have the required expertise? Who paid for the research? What is the reputation of the investigators an the institution? These are all questions that challenge

A

Credibility

82
Q

Caused by inherently unpredictable fluctuations in the readings of a measurement apparatus or in the experimenter’s interpretation of the instrumental reading

A

Random Error

83
Q

Random error can occur in either

A

Direction

84
Q

Error that is predictable, and typically constant or proportional to the true value

A

Systematic Errors

85
Q

Caused by imperfect calibration of measurement instruments or imperfect methods of observation

A

Systematic Error

86
Q

Systematic error typically occurs only in one

A

Direction

87
Q

Say you measure the mass of a ring three times using the same balance and get slightly different values of 17.46 g, 17.42 g, and 17.44 g. This is an example of

A

Random error

-can be minimized by taking more data

88
Q

Say the electronic scale you use reads 0.05 g too high for all of your measurements because it is improperly tare throughout your experiment. This is an example of?

A

Systematic error

89
Q

If the sample size is too low, the experiment will lack

A

Precision

90
Q

Time and resources will be wasted, often for minimal gain, if the sample size is

A

Too large

91
Q

Calculates how many samples are enough

A

Power analysis

92
Q

The calculation of power requires which three pieces of information?

A
  1. ) A research hypothesis
  2. ) The variability of the outcomes measured
  3. ) An estimate of the clinically relevant difference
93
Q

Will determine how many control and treatment groups are required

A

A research hypothesis

94
Q

What is the best option for showing the variability of the outcomes measured?

A

SD

95
Q

A difference between groups that is large enough to be considered important

A

Clinically relevant difference

-set as 0.8 SD

96
Q

What is the affect on sample size (n) for the following scenarios:

  1. ) More variability in the data
  2. ) Less variability in the data
  3. ) To detect small differences between groups
A
  1. ) Higher n required
  2. ) Fewer n required
  3. ) Higher n required
97
Q

What is the affect on sample size (n) for the following scenarios:

  1. ) To detect large differences between groups
  2. ) Smaller α used
  3. ) Less power (smaller β)
A
  1. ) Fewer n required
  2. ) Higher n required
  3. ) Fewer n required
98
Q

An important part of the study design

A

Statistics

99
Q

What is the null hypothesis (Ho)

A

Ho: µ1 = µ2

Ho = null hypothesis
µ1 = mean of population 1
µ2 = mean of population 2
100
Q

Is presumed true until statistical evidence in the form of a hypothesis test proves otherwise

A

Null hypothesis

101
Q

We want to compare our null hypothesis to the alternative hypothesis being tested. To do this, we must select the probability threshold, below which the null hypothesis will be rejected. This is called the

A

Significance level (α)

-Common values are 0.05 and 0.01

102
Q

Once our significance level has been selected, we need to compute from the observations the

A

Observed value (tobs) of the test statistic (T)

103
Q

Once we have calculated tobs, we need to decide whether to

A

Reject Null hypothesis in favor of alternative or not

104
Q

The incorrect rejection of a true null hypothesis (false positive)

A

Type I error

105
Q

Incorrectly retaining a false null hypothesis (false negative)

A

Type II error

106
Q

What are the two ways to compare a sample mean to a population mean?

A
  1. ) z statistic: used for large samples (n > 30)

2. ) t statistic: used for small samples (n less than 30)

107
Q

Any statistical test for which the distribution of the test statistic can be approximated by a normal distribution.

A

z statistic

108
Q

Because of the central limit theorem (CTL), many test statistics are approximately normally distributed for

A

Large samples (n > 30)

109
Q

Very similar to the z statistic and uses the same formula

A

t statistic

110
Q

When a statistic is significant, it simply means that the statistic is

-does not mean it is biologically important or interesting

A

Reliable

111
Q

Indicates strong evidence against the null hypothesis

-so we reject the null hypothesis

A

A small p-value (typically p

112
Q

Indicates weak evidence against the null hypothesis

-so we fail to reject the null hypothesis

A

A large p-value (typically p > 0.05)

113
Q

P-values close to the cutoff (0.05) are considered to be marginal (could go either way), thus we should always

A

Report our p-value so readers can draw their own conclusions

114
Q

Can strongly influence whether the means are different

A

Variability

115
Q

Most useful when comparing two means and N

A

Students t-test

116
Q

The degreesof freedom are very important in a

A

Students t-test

117
Q

Given two data sets, each characterized by it’s mean, SD, and number of samples, we can determine whether the means are significant by using a

A

t-test

118
Q

A t-test is nothing more than a

A

Signal-to-noise ration

119
Q

The degree of freedom is important in a t-test. How do we find degrees of freedom?

A

d.o.f. = N-1, but we have more than one N, so for a t-test, d.o.f. = 2N - 2

120
Q

Will test either if the mean is significantly greater than x or if the mean is significantly less than x, but not both

A

One-tailed t-test

121
Q

Provides more power to detect an effect in one direction by not testing the effect in the other direction

A

One-tailed t-test

122
Q

Will test both if the mean is significantly greater than x and if the mean is significantly less than x

A

Two-tailed t-test

123
Q

In a one-tailed t-test, the mean is considered significantly different from x if the test statistic is in either the

A

Top 5% or the bottom 5%, resulting in a p-value of less than 0.05

124
Q

In a two-tailed t-test, the mean is considered significantly different from x if the test statistic is in the

A

Top 2.5% or bottom 2.5%, resulting in a p-value less than 0.05

125
Q

If tcalc > than ttable, than we must

A

Reject the null hypothesis and conclude that the sample means are significantly different

126
Q

We must reject the null hypothesis and conclude that the sample means are significantly different if

A

tcalc > than ttable

127
Q

The observed data are from the same subject or from a matched subject and are drawn from a population with a normal distribution

A

Paired t-test

128
Q

The observed data are from two independent, random samples from a population with a normal distribution

A

Unpaired t-test

129
Q

If we are measuring glucose concentration in diabetic patients before and after insulin injection, we perform a

A

Paired t-test

130
Q

If we are measuring the glucose concentration of diabetic patients versus non-diabetic patients, we perform an

A

Unpaired t-test

131
Q

If you have more than two groups, than you must make more than two

A

Comparisons

132
Q

If you set a confidence level at 5% and do repeated t-tests on more than 2 groups, you will eventually get a

A

Type I error

-i.e. reject the null hypothesis when you should not have

133
Q

The more comparisons we have to make, the higher the

A

α value must be

134
Q

Instead of doing multiple t-tests when we have more than two means to compare, we can do an

A

Analysis of Variance (ANOVA)

135
Q

To compare three or more means, we must use an

A

Analysis of Variance (ANOVA)

136
Q

In ANOVA, we don’t actually measure variance, we measure a term called

A

“Sum of squares”

137
Q

For ANOVA, what are the three sum of squares that we need to measure?

A
  1. ) Total sum of squares
  2. ) Between-group sum of squares
  3. ) Within-group sum of squares
138
Q

Total scatter around the grand mean

A

Total sum of squares

139
Q

Total scatter of the group means with respect to the grand mean

A

Between-group sum of squares

140
Q

The scatter of the scores

A

Within-group sum of squares

141
Q

ANOVA and t-test are both essentially just

A

Signal-to-noise ratios

142
Q

To calculate the sums of squares, we first need to calculate

A
  1. ) Group means

2. ) Grand mean

143
Q

If Fcalc > Ftable,

A

We must reject the null hypothesis and conclude that the sample means are significantly different

144
Q

We must reject the null hypothesis and conclude that the sample means are significantly different if

A

Fcalc > Ftable

145
Q

When we have one measurement variable and one nominal variable, we use

A

One-Way ANOVA

146
Q

When we have one measurement variable and two nominal variables, we use

A

Two-way ANOVA

147
Q

If we measure glycogen content for multiple samples of the heart, lungs, liver, etc. We perform a

A

One-way ANOVA

148
Q

If we measure a response to three different drugs in both men and women, we use a

A

Two-way ANOVA

149
Q

Only tells us that the smallest and largest means differ from one another

A

ANOVA

150
Q

ANOVA only tells us that the smallest and largest means differ from one another, if we want to test the other means, we have to run

A

Post hoc multiple comparisons tests

151
Q

Post hoc tests are only used if the null hypothesis is

A

Rejected

152
Q

Test whether any of the group means differ significantly

A

Post hoc tests

153
Q

Don’t suffer from the same issues as performing multiple t-tets. They all apply different corrections to account for the multiple comparisons

A

Post hoc tests

154
Q

When normal distributions can not be assumed, we must consider using a

A

Non-parametric test

155
Q

Make fewer assumptions about the distribution of the data

A

Non-parametric tests

156
Q

Less powerful, meaning it is difficult to detect small diferences

A

Non-parametric tests

157
Q

Useful when the outcome variable is a rank score, one or a few variables are off-scale, or you’re sure that the data is non Gaussian (ex: response to drugs)

A

Non-parametric tests

158
Q

What is the non-parametric alternative to the two-sample t-test?

A

Mann-Whitney U test

159
Q

The Mann-Whitney U test does not use actual measurements, but rather it uses

A

Ranks of the measurements used

160
Q

In the Mann-Whitney U test, data can be ranked from

A

Highest to lowest, or lowest to highest

161
Q

Let’s say we want to test the two tailed null hypothesis that there is no difference between the heights of male and female students. What is

  1. ) Ho
  2. ) Ha
  3. ) U1
    4) U2
A

1,) male and female students are the same height

  1. ) male and female students are not the same height
  2. ) U statistic for men
  3. ) U statistic for women
162
Q

How do we analyze the Mann-Whitney U test?

A

Compare the smaller of the two U statistics to a table of U values. If Ucalc is less than U table than we reject the null hypothesis

163
Q

The extent to which two variables have a linear relationship with eachother

A

Correlation

164
Q

Useful because they can indicate a predictive relationship that can be exploited in practice

A

Correlations

165
Q

Correlation is used to understand which two things?

A
  1. ) Whether the relationship is positive or negative

2. ) The strength of the relationship

166
Q

A measure of the linear correlation between two variables, X and Y, which has a value between +1 and -!, where 1 is total positive correlation, -1 is total negative correlation, and 0 is no correlation

A

Pearson Correlation Coefficient (r)

167
Q

The goal of linear regression is to adjust the values of slope and intercept to find the line that best predicts

A

Y from X

168
Q

It’s goal is to minimize the sum of the squares of the vertical distances of the points from the line

A

Linear Regression

169
Q

Linear regression does not test whether your data are

A

Linear

170
Q

A unitless fraction between 0.0 and 1.0 that measures the goodness-of-fit of your linear regresion

-only useful in the positive direction

A

r^2

171
Q

An r^2 value of zero means that

A

Knowing X did not help you predict Y (and vice-versa)

Decks in BMS Class (62):