Final Exam - Weeks 3 & 4 Flashcards
(146 cards)
1
Q
What is bias?
A
Bias refers to any factor that systematically affects the outcomes of a study in a way that deviates from the true or expected results.
2
Q
Is bias related to study design?
A
Yes, bias can arise from various sources, such as the design of the study, the selection of participants, the measurement tools used, or the interpretation of the data.
3
Q
There is more bias in observational studies, true or false?
A
True, observational studies are more prone to bias than randomized controlled trials because they rely on the observation of existing situations and the collection of data from them, rather than the manipulation of variables in a controlled environment.
4
Q
Bias is independent of both sample size and statistical significance, true or false?
A
True, bias is related to study design.
5
Q
Explain a “spurious” finding.
A
Spurious is a term used to describe a statistical relationship between two variables that would, at first glance, appear to be causally related, but upon closer examination, only appear so by coincidence or due to the role of a third, intermediary variable.
6
Q
Explain a type I error.
A
Finding an association or significant finding when none exists.
7
Q
Explain type II error.
A
Finding no association when one truly does exist.
8
Q
Explain selection bias.
A
When the selection of study participants is not random or representative of the population being studied, leading to a distorted or inaccurate sample.
9
Q
Explain information bias.
A
When there are errors or inaccuracies in the measurement or collection of data, leading to incorrect or biased study results.
10
Q
Explain confounding bias.
A
When there is a third variable or factor that is related to both the exposure and outcome being studied, leading to a spurious or false association between the two. In other words, the observed association between the exposure and outcome may be explained by the confounding variable rather than a true causal relationship.
11
Q
Explain the general population.
A
The general population is the entire population of individuals with a characteristic of interest, such as a particular disease or condition of clinical concern.
12
Q
Explain the target population.
A
The target population of a study is the broad group of people that researchers are examining. In clinical trials the target population is often the group of people researchers believe might benefit from a particular experimental drug or treatment.
13
Q
Explain the difference between the general population and the target population.
A
The general population refers to the entire group of individuals or entities that share a common characteristic or live in a specific geographic area, while the target population is a specific subset of the general population that is the focus of a research study.
14
Q
If a study is investigating the prevalence of hypertension in the United States, the _____________________ would be all individuals who live in the country, while the _________________ would be those individuals who are at risk of developing hypertension or who have been diagnosed with the condition.
A
general population, target population.
15
Q
What is the available population?
A
The available population refers to the group of individuals or entities that are accessible and available for recruitment into a research study. This population may be a subset of the general population or the target population, depending on the study’s inclusion and exclusion criteria and the feasibility of accessing the desired population.
16
Q
The available population is always the same size as the target population, true or false?
A
False, the available population and the target population may be the same size, but this is not always true. The available population may be smaller than the target population, as not all individuals who meet the study’s criteria may be accessible or willing to participate in the study.
17
Q
Explain the sample population.
A
The sample population is a subset of the target population that is selected to participate in a research study. The sample is a smaller representation of the target population, and the selection of individuals for the sample is typically done using some form of sampling technique.
18
Q
If the sample population does not represent the target population, what type of bias is this?
A
Selection bias.
19
Q
To avoid selection bias, the sample population should be selected from the same source within the available population, true or false?
A
False! In order to avoid selection bias, the sample population should be selected from multiple sources within the available population.
20
Q
Volunteer bias, referral bias and non-response bias are all examples of ___________ bias.
A
Selection
21
Q
Recall bias, observer bias, and respondent bias are all examples of ___________ bias.
A
Information
22
Q
Use of a non-standardized or incorrect measure, and administering a measure incorrectly are examples of what type of bias?
A
Information bias.
23
Q
A ____________________ is an extraneous variable whose presence affects the variable being studied so that the results do not reflect the actual relationship between the variables under study.
A
confounder
24
Q
What is a confounder?
A
A confounder is a variable that is associated with both the exposure and the outcome in a research study, and that may distort or confound the relationship between the two.
25
In a study investigating the relationship between smoking and lung cancer, age could be a ___________________. This is because age is associated with both smoking behavior and the risk of developing lung cancer. If age is not taken into account when analyzing the data, it may appear that smoking is a stronger risk factor for lung cancer than it actually is.
Confounding variable
26
A _______________________ is associated with exposure (independent of outcome) and outcome (independent of exposure), is not in causal pathway.
confounding variable
27
If you believe a variable is a confounder, would you expect it to fit in the causal pathway?
No, confounders (confounding variables) are not in the causal pathway, they affect both the exposure and the outcome separately.
28
Related to both the exposure and the outcome BUT NOT in the causal pathway = __________________ variable.
Related to both the exposure and the outcome BUT IN the causal pathway = ____________________ variable.
Confounding
Intermediate
29
How can confounding variables be controlled for in research?
Stratification: Stratification involves dividing study participants into subgroups based on a confounding variable and then analyzing the data separately for each subgroup. This can help to control for the confounding variable and uncover any differences in the association between the exposure and the outcome.
Matching: Matching is a method of controlling for confounding by selecting study participants who are similar with respect to potential confounders. This can be done by matching on a single variable or by using multiple variables.
Study design: Researchers can control for confounding at the design stage of the study. For example, they can use randomization to ensure that the study groups are similar with respect to potential confounders.
30
You want to assess a program for improving the eating habits of shift workers. You put up flyers where many work night shifts and invite them to participate. What type of bias does this represent?
Selection bias (sample population not representative of available/target population).
31
You are conducting a multi-center study. One of your colleagues in a clinic is not comfortable using the outcome measure provided in the study protocol. They decided to use an outcome measure they have used before instead. What type of bias does this represent?
Information bias (the wrong information was collected/ information was collected using the wrong outcome measure).
32
You test 200 volunteers (100 men and 100 women), you find that lack of exercise leads to weight gain. What type of bias is represented in this case?
Confounding bias (sex was controlled for however other possible confounders were not such as age).
33
What are the four main categories of quantitative designs?
Non-experimental
Pre-experimental
Quasi-experimental
True experimental
34
Designs that do not randomize the sample, or mechanically manipulate the independent variable but may have a control group.
Non-experimental designs.
35
Survey designs, and passive observation are examples of what type of quantitative study design.
Non-experimental.
36
Explain cross-sectional study design.
A cross-sectional study design is a type of observational research study that collects data from a population or a representative subset of the population at a single point in time. The study design is called cross-sectional because it examines the relationship between the outcome of interest and the exposure or risk factors of interest at a specific point in time.
37
Cross-sectional study design is observational, true or false?
True.
38
Cross-sectional study design is an example of what type of quantitative study design?
Non-experimental (observational). The investigator does not alter the exposure status. The investigator measures the outcome and exposures in the population and may study their association.
39
___________ is the proportion of persons with a particular condition or attribute in a given population at a point in time.
Prevalence.
40
Cross-sectional study designs may be used for population-based surveys, true or false?
True.
41
Cross-sectional studies may be used for estimating the prevalence in clinic-based studies, true or false?
True.
42
Cross-sectional studies are often used in epidemiological research to estimate...
the prevalence of a disease or condition in a population, or to examine the distribution of risk factors among different populations.
43
One of the main advantages of a cross-sectional study design is _______________________________________________________________________________. However, cross-sectional studies have some limitations, such as the inability to establish causality or temporal relationships between the outcome and exposure variables.
that it is relatively quick and inexpensive to conduct.
44
Cross-sectional study design can establish cause and effect, true or false?
False, cross-sectional study design can not establish cause and effect.
45
Cross-sectional study design is unable to establish temporal relationships between the outcome and exposure variables, explain.
Establishing a temporal relationship is important in determining causality between an exposure and an outcome. In order for causality to be established, it is necessary for the exposure to precede the outcome in time. Cross-sectional study design is unable to establish temporal relationships because it assesses the exposure and outcome at the same time, without the ability to determine which event occurred first.
46
Cross-sectional research design is prone to ________________ bias.
Selection bias. Cross-sectional study designs are prone to selection bias because they rely on a sample of participants who are recruited at a single point in time. The sample may not be representative of the target population, leading to biased results.
47
What is case-control study design?
A case-control study design is a type of observational study that compares individuals with a specific disease or condition (cases) to individuals without the disease or condition (controls). The study design is retrospective, meaning that it looks back in time to identify differences in exposure to risk factors or potential causes of the disease or condition between the cases and controls.
48
Case-control study design is observational, true or false?
True. A case-control study design is a non-experimental study design, meaning that it does not involve intervention or manipulation of the exposure or risk factors.
49
Case-control study design is always prospective, true or false?
False! Case control study design is always retrospective (starts with an outcome then traces back to investigate exposure)
50
In case-control study design _____________ precedes ____________.
In case-control study design the exposure precedes the outcome.
51
Explain "case" and "control" in case-control study design?
Case refers to individuals with the outcome, condition, or event.
Control refers to individuals without the outcome, condition, or event.
52
Cases and controls need to be similar on all variables except the ____________ and the ________________.
Exposure and the outcome.
53
Cases and controls must arise from the same __________________.
general population.
54
Controls should be selected as a representative sample of the ___________________________________.
population which produced the cases.
55
Name some common sources of bias in case-control studies.
Recall bias: cases and controls may recall past exposure differently
Interviewer/observer bias: the recording of exposure information may vary depending on the investigator's knowledge of an individual's disease status.
Selection bias: controls are unrepresentative of the population that produced the cases.
56
Name some weaknesses of case-control studies.
- Ascertainment of past exposures can be difficult.
- Subject to biases.
- Selection of controls can be challenging.
- Difficult to reference records and data if record keeping is either inadequate or unreliable.
57
What type of study design is:
- Not expensive
- Efficient for studying rare diseases
- Efficient for studying diseases with a long latency period
- Able to study multiple exposures
- Good with study conditions where would be unethical to expose the participants (e.g., smoking)
Case-control study design.
58
The odds ratio (OR) is used in case-control studies to estimate...
the strength associated between exposure and outcome.
59
Explain odds ratio.
The odds ratio is a statistical measure used to quantify the strength of association between two binary variables, or outcomes. It is calculated as the ratio of the odds of the outcome occurring in one group compared to the odds of the outcome occurring in another group.
To calculate the odds ratio, first, we need to construct a 2x2 contingency table that shows the number of individuals in each group who either have or do not have the outcome of interest.
60
Explain an odds ratio of >1, <1, and =1
If the odds ratio is greater than 1, it indicates that the outcome is more likely to occur in the first group.
If the odds ratio is less than 1, it indicates that the outcome is more likely to occur in the second group.
An odds ratio of 1 indicates that there is no difference in the odds of the outcome occurring between the two groups.
61
If OR >1 and 95% CI does not include 1, then the exposure...
significantly increases odds of outcome/disease.
62
If OR<1 and 95% CI does not include 1, then the exposure....
significantly decreases odds of outcome/disease.
63
If OR=1 and 95% CI does not include 1, then exposure...
does not affect odds of disease/outcome.
64
If the 95% confidence includes 1, what does this mean?
There is no statistical significance.
65
Association between gender and fractures due to falls among patients in a rehabilitation hospital.
OR=1.71
CI=0.99, 2.97
Interpret these results.
The OR>1 meaning that the exposure increases the odds of the outcome. In this case this could mean that males had a 1.71 times higher odds or 71% higher odds of fracture due to falls compared to female participants.
The CI does not include 1 meaning that these results are not statistically significant.
66
Association between smoking and lung cancer.
OR=8.25
CI=2.52-71.37
Interpret these results.
The OR>1 meaning that the exposure increases the odds of the outcome. In this case the exposure (smoking) increases the odds of the outcome (lung cancer).
The CI does not include 1, meaning that these results are statistically significant.
67
Association between exercise and stroke.
OR=0.40
CI=0.20-0.91
Interpret these results.
The OR<1, meaning that the exposure decreases the odds of the outcome. In this case, the exposure (exercise) decreases the outcome (stroke). The findings indicate that people who exercise have 0.40 times the odds of having a stroke or have 60% lower odds of having a stroke than those who don't exercise.
The CI does not include 1, meaning that these results are statistically significant.
68
What is a cohort study design?
A cohort study is a type of observational study where a group of individuals, or a "cohort," is identified based on a shared characteristic, such as age, gender, occupation, or exposure to a particular risk factor. The cohort is then followed over time to observe the occurrence of health outcomes of interest, such as the development of a disease or the occurrence of a particular event.
69
Cohort studies can be retrospective or prospective, true or false?
True.
70
What is the best study design for determining incidence and natural history of a condition?
Cohort study.
71
In a ________________ cohort study, the cohort is identified at the beginning of the study and followed over time into the future, with data collected at regular intervals.
prospective
72
In a ___________________cohort study, the cohort is identified based on exposure status or other characteristics that have already occurred in the past, and data are collected retrospectively.
retrospective
73
What is relative risk/risk ratio?
The ratio of the probability of developing the condition if exposed to a certain variable compared with the probability if not exposed.
74
Relative risk = ____________/_______________
Relative risk= risk of outcome if exposed/ risk of outcome if unexposed
75
Association between smoking and myocardial infarction.
RR=2.45
CI=1.15-5.47
Interpret these results.
RR>1, meaning the exposure (smoking) increases the outcome (myocardial infarction). People who smoke are at 2.5 times greater risk than non-smokers of having an MI.
This finding is statistically significant because the CI does not include 1.
76
Association between regular exercise and stroke.
RR=0.85
CI=0.61-1.06
Interpret these results.
RR<1, meaning the exposure (regular exercise) decreases odds of the outcome (stroke). People who exercise regularly are at 0.85 times less risk of stroke (15% less likely).
This finding is not statistically significant because the CI includes 1.
77
Cohort study design can establish temporal sequence, true or false?
True, in a cohort study, researchers identify a group of people who are free of the outcome of interest at the beginning of the study and then follow them over time to see whether they develop the outcome. By doing so, the temporal sequence of exposure and outcome can be established, as the exposure occurs before the outcome.
78
A study is conducted to investigate the relationship between smoking and lung cancer. Researcher identified a group of people who are smokers (the exposed group) and a group of people who are non-smokers (the unexposed group). Then, they followed both groups over time and observe the incidence of lung cancer in each group. If the incidence of lung cancer is higher in the exposed group than in the unexposed group, it suggests that smoking is a risk factor for lung cancer. What type of study design is this?
Cohort study design.
79
In a _____________________ study the investigator measures the outcome and the exposure in the study participants at the same time.
Cross-sectional study
80
Participants are selected for the study based on their outcome status. The investigator then assesses the exposure in both these groups.
Case-control study
81
Participants are selected based on the exposure status of the individual. They are then followed over time to evaluate for the occurrence of the outcome of interest.
Cohort study
82
Are case-control studies or cohort studies typically more expensive?
Cohort studies are typically more expensive.
83
If the objective of the study is to determine prevalence, what study design would be the most appropriate?
Cross sectional
84
If the objective of the study is to determine incidence, what study design would be the most appropriate?
Cohort
85
If the objective of the study is to determine prognosis, what study design would be the most appropriate?
Cohort
86
Explain pre-experimental study design.
Pre-experimental designs are research designs that lack some of the key features necessary to establish a causal relationship between an independent variable and a dependent variable. They are called "pre-experimental" because they are often used as a preliminary step before conducting a more rigorous experimental design.
87
Pre-experimental designs are characterized by the absence of ___________________________ and the lack of random assignment of participants to groups. This means that researchers are unable to rule out alternative explanations for the observed relationship between the independent and dependent variables.
of a comparison or control group
88
Pre-experimental designs do not randomize the sample, but do include a control group, true or false?
False, pre-experimental designs do not randomize the sample, OR have a control group.
89
In pre-experimental designs the independent variable is manipulated, true or false?
True, the independent variable is manipulated by the researcher to observe its effects on the dependent variable. However, pre-experimental designs lack some of the key features necessary to establish a causal relationship between the independent variable and the dependent variable. This includes the absence of a comparison or control group and the lack of random assignment of participants to groups.
90
Explain one-shot design.
One-shot design is a type of pre-experimental design. This design involves measuring the dependent variable after the independent variable has occurred. There is no measure of outcome at baseline, but there is a measure of outcome post-intervention.
91
This type of pre-experimental design measures outcome post-intervention only (no measure at baseline).
One-shot design.
92
Explain static group comparison design.
Static group comparison design is a type of pre-experimental design. This design involves comparing two groups that differ in terms of the independent variable. However, the groups are not randomly assigned, which can lead to selection bias. The outcomes are compared between groups post-intervention.
93
This type of pre-experimental design measures outcome post-intervention between two groups but does not include a measure at baseline for either group. On group receives the intervention and the other does not.
Static group comparison design.
94
Explain one group pretest-posttest design.
One group pretest-posttest design is a type of pre-experimental study design. This design involves measuring the dependent variable before and after the independent variable has occurred. However, there is no comparison group, making it difficult to rule out alternative explanations for the observed changes.
95
This type of pre-experimental design compares outcome between baseline and post-intervention time points of one group.
One group pretest-posttest design.
96
_______________________ designs do not randomize the sample, but may have a control group or mechanically manipulate independent variable (an intervention).
Quasi experimental.
97
Explain quasi-experimental design.
In quasi-experimental design, the researcher manipulates the independent variable, but groups are not formed through random assignment. Instead, groups are often pre-existing or naturally occurring, and the researcher must try to control for confounding variables through various means.
98
Quasi-experimental design is a research design that shares some features with true experimental design, but lacks...
random assignment of participants to groups.
99
Quasi-experimental designs are used when...
random assignment is not possible or ethical, but the researcher still wants to establish cause-and-effect relationships between variables. They are particularly useful when the independent variable is not something that can be randomly assigned to participants, such as a person's gender, age, or medical condition.
100
Non-experimental, pre-experimental, quasi-experimental and true experimental study designs are all _______________________.
Quantitative study designs. They involve the collection of numerical data through standardized procedures and statistical analyses to investigate relationships between variables.
101
Explain non-equivalent control group design.
Non-equivalent control group design is a type of quasi-experimental design. This design involves a treatment group and a comparison group that are not formed through random assignment. Instead, the researcher selects pre-existing groups that are as similar as possible on key varibales that might affect the outcome.
102
When using a ______________study design, a researcher might compare the test scores of students in two schools that have similar demographics, but one school uses a new teaching method while the other does not without using random assignment to from the two groups.
Quasi-experimental (non-equivalent control group design).
103
In non-equivalent control group design, the outcome of both groups is measured at baseline, true or false?
True.
104
Explain some advantages of quasi-experimental study design.
- Increased external validity: often used in real-world settings and are more likely to reflect the complexity of real-life situation than laboratory experiments.
- Cost-effective: often less expensive then experimental designs, as they do not require the same level of control over extraneous variables or recruitment of large samples of participants.
- More ethical: often more ethical than experimental designs, as they do not involve random assignment of participants to groups. This can be especially important when the intervention being studied has the potential to harm participants, or when random assignment is not feasible or practical.
105
Explain some disadvantages of quasi-experimental design.
- Threats to internal validity: susceptible to threats to internal validity, such as selection bias, history effects and maturation effects. This means that it is more difficult to establish a causal relationship between the independent and dependent variables.
- Lack of randomization: do not involve random assignment of participants to groups, which can increase the risk of confounding variables affecting the results.
- Limited control over extraneous variables: making it more difficult to attribute any observed effects solely to the manipulation of the independent variable.
106
Why is there a risk of confounding bias in quasi-experimental study design?
Confounding variables are usually balanced by the randomization process.
107
This type of quantitative study design randomizes the sample, has a control group and mechanically manipulates the independent variable.
True experimental.
108
Explain true experimental study design.
True experimental study design is a research method used to establish a cause-and-effect- relationship between two or more variables. In this design, the researcher manipulates an independent variable, while controlling for other variables, and observes the effect of this manipulation on a dependent variable.
109
In this type of quantitative study design, the researcher manipulates an independent variable, while controlling for other variables, and observes the effect of this manipulation on a dependent variable. The design involves random assignment of participants to groups to ensure that any observed effects are due solely to the manipulation of the independent variable, rather than any pre-existing differences between the groups.
True experimental study design.
110
Randomized controlled trials, posttest only design, Solomon four-group designs, factorial designs and counterbalance designs are all what type of study design?
Quantitative (true experimental).
111
The ________________________ design is considered the gold standard in research because it provides the strongest evidence for causality between the independent and dependent variables. However, this design can be challenging to implement in some research settings, and may require a large sample size and significant resources to carry out.
True experimental.
112
Which true experimental study design do these characteristics relate to?
- The intent of the research is to evaluate a deliberate intervention which is often an innovation in treatment.
- Intervention and comparison groups are deliberately manipulated by the investigator.
- Groups are assigned by a random process.
- Groups enter at the same calendar time and are followed over the same time period.
- The groups are compared.
Randomized controlled trial.
113
Name some advantages of RCTs.
- Control over extraneous variables: random assignment
- Causality: can establish causality by demonstrating a statistically significant difference between treatment and control groups. By randomizing participants and using a control group, the design can demonstrate that the treatment is responsible for observed changes.
- Reliability: use standardized methods and protocols for data collection and analysis.
- Generalizability: to a larger population.
- Acceptance in medical practice: RCTs are widely accepted in medical practice and research.
- Strongest evidence for a causal relationship between an exposure and an outcome.
114
Name some challenges associated with RCT study design.
- Often complex and time-consuming
- Expensive
- Require a large sample size, struct protocols, and well-trained staff
- RCTs often involve a placebo group which can be ethically challenging as it involves withholding treatment from some participants.
- Attrition or loss of participants during the study can affect the power of the study and reduce the generalizability of the results.
- Limited external validity: while RCTs can establish causality, they may not always be applicable to real-world settings. The strict protocols and highly controlled environment of RCTs may not always reflect real-world settings.
115
RCTs are intended to answer questions about the ______________ or ______________ of an intervention.
efficacy or effectiveness
116
Why would we use PICOT format not PECOT for RCT research questions?
PICOT is used over PECOT because RCTs include an intervention (I).
IN PECOT the E stands for Exposure: the intervention, exposure or risk factor being studied.
117
PECOT is typically used in ________________________ studies, where the exposure or risk factor being studied may not necessarily be an intervention or treatment. PICOT, on the other hand, is typically used in ______________________ studies, where the focus is on a specific intervention or treatment.
Observational
Interventional
118
The _____________________ refers to a phenomenon in which study participants modify their behavior in response to being observed or studied, rather than the intervention or treatment being tested. This can result in changes in behavior or outcomes that are not due to the intervention or treatment itself, but rather to the participants' awareness of being studied.
Hawthorne effect.
119
Strongest evidence for causality comes from a design where subjects are assigned ______________ to the interventions.
at random
120
Explain the advantages of randomization.
- Probability of group assignment is known and equal for all (does not depend upon characteristic s of subjects)
- Limited selection bias: subjects are not assigned (or chose) intervention based on propensity to have a good outcome.
- Characteristics of subjects, known and unknown, are assigned along with the subject
- Limited confounding bias: on average, in larger sample sizes, randomization balances group on confounding factors (know and unknown)
- Limited information bias: as long as outcome assessors are blind to group assignment
121
Randomization only guarantees balance at the outset of the trial, explain why.
As time goes on, subjects may drop out of trial, they may cross-over (to the intervention or away from it). If these individuals are excluded from the final analysis, the only people who remain are those who selected themselves to remain where they were assigned (selection bias). A solution to this bias is to ensure that for the analysis, everyone must stay in the groups to which they were assigned, and to report subjects path through the study (CONSORT diagram).
122
It is not always feasible to blind the subjects and assessors based on the nature of the intervention, name a possible solution for this dilemma.
If possible choose an outcome not affected by observer bias, for example the 2-min timed-up-and-go test is an assessment that could be used as an outcome free of observer bias.
123
Explain single and double blinding strategies in research study design and explain the difference between the two in effectiveness of reducing bias.
Single blinding involves masking the participants to the treatment they are receiving, while the investigators are aware of the treatment assignments.
Double blinding involves masking both the participants and the investigators to the treatment assignments.
Double blinding is generally considered to be more rigorous and effective at reducing bias compared to single blinding. This is because double blinding eliminates the potential for bias due to both the participant and the investigator knowing the treatment assignments, and it reduces the potential for placebo effects or other factors that could influence the study outcomes.
However, double blinding may not always be feasible or necessary in certain types of studies. For example, in some studies, it may be difficult or impossible to blind the participants, such as in surgical interventions. In other studies, blinding may not be necessary or may not be possible due to the nature of the intervention or treatment being studied.
124
What is effect size?
Effect size refers to the magnitude or strength of a statistical relationship or difference between two variables. It provides a standardized measure of the size of an effect, allowing researchers to compare the strength of effects across different studies or experimental conditions. (magnitude of the effect due to the intervention)
125
Smaller effect size require _______________ samples to detect.
Larger samples.
126
Interpret the effect size of the following:
0.0 to 0.2 =
0.2 to 0.5 =
0.5 to 0.8 =
0.8 and above =
0.0-0.2 = trivial effect size
0.2-0.5 = small effect size
0.5-0.8 = moderate effect size
0.8 and above = strong effect size
127
While_____________________ indicate whether an effect is statistically significant (i.e., unlikely to have occurred by chance), ___________________indicates how large the effect is in practical terms.
p-values
effect size
128
For example, a ____________ effect size may be statistically significant in a large sample size, but it may not be clinically relevant. On the other hand, a ___________ effect size may be clinically relevant even if it is not statistically significant, particularly if the sample size is small or the effect is difficult to measure accurately.
Small
Large
129
Explain why a trial may be stopped early if an effect size occurs.
A trial may be stopped early if the effect size is very large, which means that the intervention being tested is having a much greater impact than expected. Stopping a trial early due to a large effect size is often done for ethical reasons, as it allows participants in the control group to receive the intervention that is proving to be effective.
For example, if a trial is testing a new drug for a life-threatening disease and the effect size is very large, it may be deemed unethical to continue the trial with a control group receiving a placebo, as it would deny them the potentially life-saving treatment. In such cases, the trial may be stopped early, and the treatment may be made available to all participants in the trial.
130
Explain possible reasons for an effect size having an influence on a decision to stop a trial.
- Unexpected harmful effects
- Unexpected beneficial effects
- Futility in that no effect is likely ever to be observed
- Data quality issues that need to be corrected if the trial is to continue
131
A Consort diagram is a useful tool to review for detecting...
Selection bias
132
P-value tells you...
if the results produced by the intervention are statistically significant when compared to the results produced by the control group.
133
Explain the null hypothesis
The null hypothesis is a statement that there is no statistically significant difference between two groups, or no statistically significant relationship between two variables.
134
It is a hypothesis of no effect or no difference.
Null hypothesis.
135
In a study comparing two treatments for a particular disease, the ______________________ would be that there is no statistically significant difference in outcomes between the two treatments.
Null hypothesis.
136
Explain the alternative hypothesis.
The alternative hypothesis is a statement that contradicts the null hypothesis and suggests that there is a statistically significant difference or relationship between two groups or variables being compared.
137
It is the opposite of the null hypothesis and is often denoted as H1.
Alternative hypothesis.
138
When the p-value is lesser than or equal to alpha level...
When the p-value is greater than or equal to alpha level...
When the p-value is lesser than or equal to the p-value, we reject the null hypothesis (the groups are different and this difference is statistically significant)
When the p-value is greater than or equal to the p-value, we fail to reject the null hypothesis (we accept the null hypothesis) (the results are not statistically significant)
139
Explain the alpha level.
The alpha level, also known as the significance level, is the probability level used in hypothesis testing to determine whether the null hypothesis should be rejected or not. It is denoted by the symbol α and is usually set at 0.05, 0.01, or 0.001.
140
Confidence interval = _________________
CI = 1 - alpha level
141
For RCTs the null value = _____
0
142
The result would not be statistically significant if the confidence interval includes the _____________.
Null value.
143
Designs that do not randomize the sample, or mechanically manipulate independent variable but may have a control group.
Non-experimental designs.
144
Surveys are a type of __________________ study design, when used to make comparisons at a single point in time.
cross-sectional
145
Cross-sectional study designs are __________________(what type of broader study design).
observational.
146
We are interested to know the prevalence of ADHD among students in primary schools of city of X. We design a population-based survey to assess the prevalence of this condition. We go to all the schools that were supposed to be included in the study and examine the population. What type of study design is this?
cross-sectional.
