"If you don't know the cell types and the connections, how do you really understand how the machine works?"
Dr. Andrew Huberman
"Studying female rat sexual behavior in a cage is like trying to study swimming behavior of a dolphin in a bathtub."
"We don't want to use all of our data; we want to split it."
"We have to say, what are my own biases as a scholar, as a researcher, and how do I step back from those and approach it in a kind of a neutral and even-handed way?"
"This data-driven approach requires input from a wide variety of sources."
"You might be thinking, well why if you're interested in the shape, would you bother using color to pick those out? Why wouldn't you just eyeball all the images?"
"In science, we work to disprove the null hypothesis. If we fail to disprove the null, then the null is accepted to be true."
"There's almost a desired p-value for p-values."
"We can say whether these two fragments belong to the same individual, if they do, the matching that they did is okay."
"The evidence is really what dictates where you go when you research and your discoveries are sometimes serendipitous."
"We can't generally sample the whole population, so it's important to get a representative sample to avoid biases."
"Representative sample: it's about how representative it is in order to get that accuracy you need to truly inform accurate decision making."
"You choose your plot type based on the question you're answering and the data types you're working with."
"Methodology is really a philosophical stance on research."
"If you want to get the answer to the question you have to ask the right question and then you have to design the study appropriately to answer that question."
"A good scientist has to be able to switch between being thrilled at the identification of a new idea and then has to turn around and throw everything at the process of trying to falsify it."
"Be a critical thinker and look at the totality of the topic, just don't look at one little part of it."
"A fact doesn't change, but Darwin's theory is a research methodology—it's incomplete, it will change over time."
"We need good confirmed data, and when you have consistent witness testimony that presents an identifiable pattern which this does, I feel it is irresponsible not to use it."
"Most ground-based telescopes will have a proposal process and there will probably be a fair amount of time on the telescope that goes to individual small projects."
"Two separate data sources from two separate entities."
"The more methodologically accurate and stringent the research is, the less likely it is to find evidence for the hypothesis that more guns means more crime."
"More tests, more samples offer a statistically better chance of success."
"Balance the research but then also do a gut check."
"We're doing this so we can eliminate variables from the surrounding environment."
"The entire basis of academic research... is fundamentally just looking at the parts."
"If we assume the alternative hypothesis is true, we would reject the null hypothesis because it's definitely not within the non-rejection region."
"So we propose two hypothesis in the paper one is that perhaps the shell became stuck in the throat choked the fish it drowned and was preserved like this."
"David Paulides put together his profile points to understand these cases better."
"Selection bias is the bias introduced by selecting individuals that aren't necessarily representative of the entire population. This can often lead to false conclusions being drawn."
"We wanted to know if the substance was the same for all the samples."
"You gotta treat it like a scientific experiment."
"It's not really about sources, it's about methodology."
"The actual process of your work sounds very much like a Sherlock Holmes with a PhD trying to piece together stuff."
"Wrapping this one up, I want to highlight that all of my research is conducted through cross-examination from public sources."
"It's funny, 35 years ago a group of archaeologists came here, dug a load of trenches, got some finds, and then joined the dots together and conjectured that there was a Roman bath house here."
"How far out from the mean our results might be and whether that deviation from what we expect is due to chance alone or something else."
"Science has never been... without an RCT you cannot trust."
"Polls matter, but the methodology matters as well."
"One-way ANOVA tests differences between more than two groups."
"It's really important for us to check and be current on what those assumptions might be because it might typically be more than just normality."
"One thing that can be useful is to look at other published studies that have used similar designs and then look at what analysis technique they used."
"The best way to minimize selection bias is to have a representative population of the target population that you actually want to apply the results on."
"So in the documentary they were saying you would never get rid of your notes well you would never get rid of your findings you would never get rid of what you discovered but you might discard your temporary notes in that process."
"Your methodology justifies the tools in which you're using so think about a plumber a plumber receives a call to go out to a property to fix a leaky pipe or a leaky radiator the plumber will attend with a box of tools they are your research tools."
"In a retrospective cohort study, now outcomes have also occurred and of course exposure happened in the past. We are ignoring the outcome status. We are going back to the past. This is how it is different than a case control study design."
"Which study design are you going to choose? Case control or cohort study design? It will depend on the nature of the disease under investigation, the type of exposure, and the available resources."
"In a case control study, we identify a group of individuals who have the outcome of interest and we call them as cases. And we also identify a group of individuals who do not have the outcome of interest and we call them controls."
"Randomized control trials are the gold standard for drawing causal inferences."
"Researchers must examine themselves in autoethnography to understand biases."
"Randomization and double-blinding are methods that avoid biases, elements which can influence the measurement and the efficacy and lead to a wrong result."
"The most important thing about a statistical method is not how you analyze the data but what data you include."
"You always have to think about the data source based on a research question."
"You'll get the best data if you do all your experiments on the same day in the same session."
"If you choose the wrong methodology, it can cost you between 18 months and 36 months of your life."
"Confirmatory factor analysis can be your second step if you perform EFA first."
"It minimally organizes and describes your data set in rich detail."
"Systematic reviews should follow a set protocol which clearly sets out the review question, the inclusion criteria for randomized control trials, and the data collection process and procedure in an explicit and transparent way."
"The purpose of the meta-analysis is to combine the result and try to find out the effect size."
"The key to a systematic approach in social research is the idea of design."
"Simple random sampling involves selecting participants in a completely random fashion where each participant has an equal chance of being selected."
"They just observed data and they did it longitudinally, that means over a long time course with a specific selection of people."
"The methodology... that's the thing that gives a study credibility."
"Studying these things are complicated and difficult, this is why researchers try to isolate all these other factors to be able to find the causation, not just a correlation."
"Systematic review papers are reproducible."
"We look at the totality of evidence, but then you need a system to rank it."
"Research design influences research results and conclusions."
"Grounded Theory builds theories inductively based on observed patterns of events or behaviors."
"We used a method of research by design, which I think is a very important method for architects."
"Difference in differences is a method that identifies the impact by considering two differences: the difference between the treatment group and the control group, and the difference before the intervention and after the intervention."
"Remember, research builds on itself and what has been published and accepted as credible research is a good starting point for figuring out where you want to go with your research."
"The methodology chapter is not just there to say what you did, but it is so that other people could repeat it."
"Broad questions are good because it increases the applicability of the results and facilitates detection of bias."
"We need specific evidence showing the causation that that hypothesis is proposing, the causal order, chronology is very important."
"The advantage of this new framework is to facilitate a comparison across studies and to also monitor patient outcomes."
"The search methods must be transparent and you must be able to reproduce your search."
"This analysis allows me to keep every single item a participant saw so that maybe hundred to three hundred data points instead of reducing them down to one."
"The first step that I'm going to take is I need to actually be able to import data for both groups."
"This is an example where you could be reasonably confident that there's no evidence of publication bias because you've got a nice symmetrical funnel."
"Primary data is first-hand data that you collect yourself."
"Research design is to do with collecting information about the world in some kind of structured and socially approved manner."
"The only study that gives us cause and effect between two variables is experimental design."
"A random sample is a sample in which each item from the population is equally likely to be drawn."
"I try to be very evidence-based in my research, but I don't discount the value of anecdotal evidence that's out there."
"I think the way that DAGs push us to clarify our ideas is actually one of the great benefits of DAGs."
"The tagging process is supposed to be pretty painless for the bass."
"Always drill into the original source."
"There are three most important types of biases in sampling: selection bias, undercoverage bias, and survivorship bias."
"To reduce interpolation bias and the risk of overfitting, we are going to restrict typically the donor pool to units that are similar to the treated units."
"First step to know how to analyze and conduct bibliometrics, you have to read the foundational paper on bibliometrics."
"A good lit review is an argument, it's answering a question and that's an argument."
"When a researcher asks a question, he wants to know that the answer he's getting is the answer to that question."
"The strength of the panel survey is judging change over time."
"Every respondent gets an equal chance of selection; that is the beauty of probabilistic sampling where the biasness is not there."
"The accuracy improves if we have more data points to work with."
"A useful next step after performing principal components analysis is to apply cluster analysis to your data projected onto the principal components."
"The whole point of scientific research is trying to prove yourself wrong."
"PICO stands for Population, Intervention, Comparison, Outcome, and a silent T for Time frame."
"Multi-regional clinical trials readily permit structured evaluations of regional consistency of results across sub-populations of patients."
"A sample should represent all of the elements of the population as a whole."
"So now what we can do is we can now go ahead and copy all of the values from the DE tables, paste them on an Excel spreadsheet, and organize and analyze all of the values from there."
"A good idea to balance your panel is to ensure that each cross-sectional unit appears in every period in the panel."
"A structural equation modeling framework is an extremely powerful and flexible multivariate methodology."
"When we do placebos, neither the person giving nor the person receiving knows, which one's the right drug and which one's just an inert substance."
"Research is all about educated guesses, and that's the best we can do is a really good educated guess."
"If we do find effects, we're going to find a lot of smaller effects, and so we're going to need bigger studies and more collaboration."
"Interventional studies need to be randomized, they need to be blinded, and they need to be controlled."
"Every step that we take in the PRISMA checklist or the PRISMA methodology helps us to remove or reduce bias in our studies."
"Historians establish what probably happened in the past on the basis of probabilities."
"Causal research is a cause-and-effect study."
"Quantitative research collects data that can be explained numerically; it's used to determine general trends."
"Every X amount of people." - Systematic Sampling Method
"Remember there is no generalization if you are going for the non-probability sampling."
"Don't go with the EFA in this scenario; you just simply need to go with a CFA."
"Everything we ran through here is not focused on the calculations; those are of secondary importance compared to understanding the concept of what is power, what affects it, how can we control it."
"The permutation approach is really the way that group analyses are now done for assessing significance."
"Everything we did was to try to build the test distribution that was a test statistic which was pivotal."
"We always compare two groups or two chunks."
"Control group is very, very important."
"This has eliminated the chance that we see a difference in treatment effect that is not actually due to the treatment."
"We must give names to the different parts, so we know what we are referring to in our study."
"Analyze the data when collected; there's no need to collect superfluous data."
"Differentiate between input data and output or performance data."
"...be sharp with your theory and be sharp with your method."
"We experimented gathering data and answering the questions when they came up was better."