"The incidence of breast cancer in Seoul's Gangnam area and Seongnam city's Bundang district in Gyeonggi Province has been found to be higher than in other areas. According to the 'Cancer Incidence Statistics by City and County' first announced by the Ministry of Health and Welfare and the National Cancer Center on the 22nd, the area with the highest occurrence of breast cancer from 2009 to 2013 was Seoul's Seocho district, with 65.1 cases per 100,000 people. This was followed by Seoul's Gangnam district (64.4 cases), Gyeonggi Province's Yongin city's Suji district (63 cases), Seongnam city's Bundang district (62.2 cases), and Busan's Gangseo district (62.1 cases). Considering that the national average during the same period was 49.5, these numbers are quite high."
- https://www.hankookilbo.com/News/Read/201611221717117488
Getting rich raises the risk of cancer. Maybe being greedy cause cancer?
Contents
1. Correlation and Causation
2. Only experimental study can claim causation
3. Claiming causality without experimental study
4. Qausi-experiment cannot claim Causation
1. Correlation and Causation
Correlation: Suppose you observe that during the summer, both ice cream sales and the rate of sunburns increase. You could say there is a positive correlation between ice cream sales and sunburns. That is, as ice cream sales go up, so do sunburns, and vice versa.
However, correlation does not imply causation.
Causation: It would be incorrect to conclude from the correlation that eating ice cream causes sunburn (or that getting a sunburn causes one to buy ice cream). Why? Because there's a missing variable here: the weather. In the summer, the weather gets warmer and sunnier. Warm, sunny weather is the underlying cause of both increased ice cream sales (people tend to eat more ice cream when it's hot) and sunburns (people tend to get more sunburns when they're exposed to more sunlight).
So while ice cream sales and sunburns are correlated, one does not cause the other. Instead, both are caused by a third variable: the weather.
This illustrates the important concept that while correlation can indicate a relationship between two variables, it doesn't necessarily mean that one causes the other.
2. Only experimental study can claim causation
The core difference between observational studies and experimental studies lies in how the data is gathered and how much control the researchers have over the variables in the study.
Observational Studies: In these studies, researchers observe subjects in their natural settings and do not intervene or manipulate any variables. They merely record what happens. For instance, in an observational study of smoking and lung cancer, a researcher might record the number of cigarettes smoked and the incidence of lung cancer, but they would not control who smokes or how much they smoke. These studies can reveal correlations or associations, but they can't definitively prove causation because there may be confounding variables at play that the researchers are not controlling for. For example, maybe smokers in the study also tend to have poor diets or live in polluted areas, and those factors could be contributing to the increased incidence of lung cancer.
Experimental Studies: In contrast, experimental studies involve the manipulation of one or more variables to see how this affects other variables. Researchers assign subjects to different conditions and control all other factors as closely as possible. For instance, in an experimental study on a new drug for hypertension, researchers might randomly assign half the subjects to take the drug and half to take a placebo, while controlling for variables such as age, gender, diet, and exercise. Because of this control and manipulation, researchers can make stronger inferences about causation from experimental studies. If the group taking the drug shows a significant reduction in hypertension compared to the placebo group, the researcher can infer that the drug causes a reduction in hypertension.
Experimental studies can provide evidence of causation because they can rule out confounding factors through the random assignment of subjects to conditions and the careful control of other variables. This allows researchers to say with greater confidence that the changes they observe in the outcome variable (like hypertension) are due to the manipulation of the independent variable (like the drug), and not due to some other factor they weren't controlling for.
2-1. Herbal tea and sleep quality.
Correlation:
Suppose a researcher conducts a survey where they ask a large group of people about their tea-drinking habits and their sleep quality. The researcher notices that people who reported drinking chamomile tea at night also reported having better sleep quality.
This is an example of a correlation – people who drink chamomile tea tend to have better sleep. However, this doesn't mean that chamomile tea causes better sleep. There could be other variables at play. For example, those who drink chamomile tea might also have a nightly routine that promotes good sleep, like reading a book or turning off electronic devices an hour before bed.
Causation:
To prove causation, the researcher could conduct an experiment. They might divide participants into two groups: one group is given chamomile tea to drink each night, and the other is given a placebo tea. If, after a few weeks, the chamomile tea group reports significantly better sleep than the placebo group, the researcher could make a stronger claim that chamomile tea causes improved sleep.
It's important to note that even in this controlled experiment, the finding does not absolutely confirm causation. There could still be other factors involved. However, the experimental design does provide a more solid basis for suggesting that drinking chamomile tea improves sleep quality.
2-2. Tai Chi practice and hypertension.
Correlation in Observational Studies
In an observational study, researchers merely observe the subjects without intervening. They then gather data and attempt to identify patterns and relationships.
For instance, suppose you're interested in whether daily Tai Chi practice affects hypertension, an observational study might involve finding a large sample of people who practice Tai Chi daily and a similar group who do not. You'd then compare the rates of hypertension in these two groups.
If it were observed that the Tai Chi group had significantly lower rates of hypertension, we could say there is a correlation between Tai Chi practice and lower hypertension. However, because it's an observational study, we can't say for certain whether the Tai Chi is the cause of the lower hypertension. It might be that people who choose to practice Tai Chi daily also tend to eat healthier diets or engage in other behaviors that reduce hypertension.
Causation in Experimental Studies
In an experimental study, the researcher manipulates one variable to see how it affects another variable. This approach allows us to make stronger inferences about causality.
For example, to determine whether Tai Chi practice causes a reduction in hypertension, you might conduct a randomized controlled trial (RCT). In this study, you would randomly assign a large group of people with hypertension to either a treatment group (who would be instructed to practice Tai Chi daily) or a control group (who would maintain their normal daily routines).
If after a certain period (let's say six months), the treatment group shows a statistically significant reduction in hypertension compared to the control group, we could infer a causal relationship. That is, we could say that daily Tai Chi practice leads to a reduction in hypertension. This conclusion comes with more confidence because the random assignment of participants to the treatment or control group helps to ensure that the observed effect is not due to other confounding variables.
2-3. Finishing the test early and getting good scores
Every final term, I observe a pattern that students who finish a final test earlier tend to get higher scores. So, I encourage my students to submit the paper as soon as possible.
'Finishing the test early and getting good scores' is a correlation -- there's a relationship or association between the two variables (time taken to finish the test and the test score). But it's crucial to remember the principle "correlation does not imply causation." We cannot immediately conclude that finishing the test faster causes higher scores based on this observed correlation alone. Claiming causation would be saying that finishing an exam earlier directly causes higher exam scores. This is a stronger claim and requires evidence that rules out other potential explanations.
There might be several other factors at play that could explain this correlation:
Preparation: It's possible that students who are better prepared for the test can answer the questions faster and also tend to get higher scores because of their preparation, not because they finished the test early.
Test-taking skills: Some students might be more skilled at test-taking strategies, such as time management or quickly understanding and answering questions, which could lead them to finish earlier and score higher.
To test for a causal relationship, you could conduct an experimental study in which students are randomly assigned to take the test under different time conditions.2-4. G
2-4. Getting risk raises the risk of cancer.
Can you explain why this is not fair statement based on prior study introduced?
Correlation:
An observational study found a correlation that certain areas, such as Seoul's Gangnam district and Seongnam's Bundang district, have higher rates of breast cancer than other regions. This means that living in these particular areas and higher breast cancer rates appear to go together more frequently than one would expect by chance.
However, this does not mean that living in these areas causes breast cancer. There could be various confounding factors at play. Perhaps these areas have better access to healthcare services, and thus breast cancer is detected and reported more frequently. Or, there might be lifestyle or environmental factors unique to these regions that contribute to the higher rates.
Causation:
To establish a causal relationship between living in these areas and the higher rates of breast cancer, more rigorous studies would be needed. For instance, a researcher could conduct a controlled experiment in which they compare women of the same age and similar health status who live in these areas with those who live in areas with lower breast cancer rates, while controlling for other factors such as healthcare access and lifestyle habits. If the breast cancer rate remains significantly higher in the Gangnam and Bundang districts after controlling for these factors, it could suggest a causal relationship.
However, establishing causation in this context would be difficult, largely due to the multitude of potential variables and the complexity of conducting such controlled experiments in real-world settings. As always, correlation does not imply causation, and it's important to thoroughly investigate all potential factors when considering these kinds of relationships.
It can raise another question, then how can you prove getting rich raise the risk of cancer?
2-5. Conundrum: Experiment and Ethics
Can you design the experiment which can prove the fact 'Barking dogs never bite'?
Hypothetical experiment that could be designed to investigate the relationship between a dog's ability to bark (independent variable) and biting incidents (dependent variable). But remember, in real-world research, animal welfare and ethical guidelines should always be of primary importance, and the experiment described below would be ethically inappropriate to conduct.
Step 1: Selection of Participants
Acquire a large, diverse sample of dogs of varying breeds, sizes, and ages.
Step 2: Pre-Experimental Observations
First, record baseline behaviors of all dogs over a set period of time in a neutral setting. Note any barking or biting incidents.
Step 3: Randomization
Randomly assign the dogs to two groups: the 'Barking' group and the 'No Barking' group. For the sake of this hypothetical, let's say the 'No Barking' group have their vocal cords temporarily and harmlessly impaired through a reversible method, such as using a safe, temporary numbing agent. (Again, it is essential to remember that this would be ethically unacceptable in reality.)
Step 4: Experimental Conditions
Expose both groups to a variety of stimuli known to provoke barking in dogs. These stimuli could include a knock on the door, the presence of an unfamiliar person, or the sound of a doorbell. Monitor the dogs closely throughout this period.
Step 5: Measurement
Record all incidents of barking and biting for both groups. The researchers would need to carefully observe and record the behaviors of the dogs in response to the stimuli.
Step 6: Data Analysis
Compare the incidence of biting in the 'Barking' group versus the 'No Barking' group. If the dogs in the 'No Barking' group exhibited more biting behaviors, you might conclude that there is a relationship between barking and biting.
Ethical guidelines for scientific research, particularly involving animals, require that any harm or discomfort caused to subjects be minimized as much as possible and be justified by the potential benefits of the research. These guidelines exist to protect the welfare of research subjects and to ensure that the research is conducted responsibly.
Institutional Review Board" (IRB) / "Ethics Committee for Clinical Trials.
An IRB is an independent committee responsible for reviewing and ensuring the ethical and legal conduct of clinical trials and other human research studies. Its primary role is to protect the rights, welfare, and well-being of the participants involved in research.
The IRB evaluates research protocols, study designs, and related documents to ensure that the study is scientifically valid and ethically sound. It assesses the potential risks and benefits of the research, considers the informed consent process, and ensures the inclusion of vulnerable populations, such as children or individuals with impaired decision-making capacity. The committee also monitors ongoing studies to ensure compliance with ethical standards and regulations.
The IRB consists of a diverse group of professionals, including medical experts, scientists, ethicists, legal advisors, and community representatives. Their collective expertise helps ensure that research studies are conducted in a responsible and ethical manner, following established guidelines and regulations.
By overseeing the ethical aspects of clinical trials, the Institutional Review Board plays a crucial role in safeguarding the rights and well-being of research participants and maintaining public trust in the scientific community.
*The IRB reviews research that involves human participants. Research involving animals must be reviewed by Committee for Animal Care (CAC).
3. Claiming causality without experimental study
How can you claim smoking induce lung cancer without experimental study?
Due to ethical reasons, there have not been experimental studies that randomly assign individuals to either smoke or not smoke to see who develops lung cancer. Such a study would be highly unethical as it could knowingly cause harm to the subjects.
In the case of smoking and lung cancer, the strength of the evidence from observational studies is so strong that the scientific community has reached a consensus that smoking does cause lung cancer. There's a wealth of observational epidemiological evidence that establishes a strong link between smoking and lung cancer. This relationship has been consistently observed across different populations, time periods, and study designs. These are the factors need to be considered as storng evidence;
- Consistency of the association: Numerous studies have consistently found that smokers are much more likely to develop lung cancer than non-smokers.
- Strength of the association: The risk of lung cancer increases with the amount of smoking, and decreases when smoking is stopped, suggesting a dose-response relationship.
- Specificity: Lung cancer rates are significantly higher among smokers than among non-smokers, suggesting a specific link.
- Coherence: The association is coherent with what we know about the harmful substances in tobacco smoke and how they can damage lung tissue and lead to cancer.
- Temporal relationship: The fact that lung cancer usually develops after many years of smoking supports the idea that smoking can cause lung cancer.
Although it's true that correlation does not imply causation, when observational studies demonstrate a strong, consistent, and specific association, and there is a plausible mechanism that explains the relationship, it's reasonable to infer causation. The evidence linking smoking to lung cancer meets these criteria.
4. Qausi-experiment cannot claim Causation
A 'before and after' study, also known as a pretest-posttest design, is often considered an observational study or a quasi-experimental study, as it does not include a control group for comparison. It investigates the changes in an outcome (dependent variable) before and after a treatment or intervention (independent variable) in the same group.
For example, a 'before and after' study might measure individuals' blood pressure, administer an acupuncture treatment, and then measure blood pressure again to see if it has changed. If blood pressure is lower after the acupuncture treatment, researchers might conclude that the acupuncture had an effect.
While this type of study can provide evidence of change over time, it can't definitively prove causation for several reasons:
Lack of control group: Without a control group that did not receive the acupuncture treatment, it's difficult to say whether the observed changes in blood pressure were due to the treatment or to some other factor.
Natural change over time: Some changes could be due to the passage of time, not the intervention. For instance, people might naturally relax during the course of the study, causing their blood pressure to drop.
Placebo effect: People might expect the acupuncture to lower their blood pressure, and this expectation could influence their actual blood pressure.
In contrast, a true experimental design would include random assignment to a treatment group that receives acupuncture and a control group that does not. This design allows for a more direct comparison and helps control for other potential variables, making it more likely to accurately determine if there's a causal relationship between acupuncture and blood pressure reduction.
Read the article: How to tell experimental study
https://acupunctureherbalmedicine.blogspot.com/2023/05/comparing-effectiveness-of-pickup-lines.html
Final Verdict
The effect of acupuncture at the Yuji point on resting-state brain function in anxiety
Background: The COVID-19 epidemic has placed a lot of mental burdens on school students, causing anxiety. Clinically, it has been found that the Yuji point (LU10) can relieve anxiety by regulating Qi.
Methods: Thirty-six volunteers with anxiety disorders were divided into 3 groups, all of whom underwent 2 MRI examinations. The Yuji and nonacupoint groups received acupuncture between functional magnetic resonance imagings. We used the amplitude of low-frequency fluctuation to analyze regional brain activity, and seed-based functional connectivity (FC) to analyze changes in brain networks.
Results: After acupuncture, the LU10 was able to activate the frontal lobe, medial frontal gyrus, anterior cingulate gyrus, temporal lobe, hippocampus, etc in the left brain compared to the control group. The frontal lobe, medial frontal gyrus, cingulate gyrus, and anterior cingulate gyrus in the left brain were activated compared to those in the nonacupoint group. Compared with the control group, LU10 showed increased FC in the right parietal lobe, right precuneus, left temporal lobe, left superior temporal gyrus, and with cingulate gyrus. FC was enhanced among the hippocampus with the left temporal lobe and the superior temporal gyrus and reduced in the right lingual gyrus and right occipital lobe.
Conclusion: Acupuncture at LU10s can regulate anxiety by upregulating or downregulating the relevant brain regions and networks. LU10s can be used to treat not only lung disorders but also related mental disorders.