Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Molecular Biology > Experimental Design > Flashcards

Experimental Design Flashcards

You may prefer our related Brainscape-certified flashcards:
AP® Biology flashcards Biology 101 flashcards
1
Q

What is Science?

A

-Making predictions that allow you to set up conditions that you then see the outcome and the science will determine the outcome (no influence from you on the results)

  • Conclusions = need to be careful with forming one because they can lead to bias (e.g., wanting the results to be similar to previous experiments even if it’s not)
  • Question –> Predict –> Experiment
    Observe (keep good notes of your observations)
    Analyze (use logic to make a conclusion and think about comparisons)
    Report ( need to otherwise no one will know what you did or learn from it and it becomes a waste of resources) = publications/presentations
    -formulate new questions/hypothesis (ongoing work)

-Humans are the weakest link because we can have error in data interpretation

-Discussion of limitations is highly important as well

  • ability to disprove, you can never prove anything
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Behaving Scientifically

A
  1. understand background and context
  2. Think about what you missed (limitations, gaps)
  3. Do not ignore evidence you don’t like (if you get data you don’t like, find another way of asking the same question)
  4. Use all the data
  5. communicate
  6. share
  7. act with integrity, ethics
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Edward Tolman’s work in psychology was is science?

A
  • wanted to explain rat brains (=nature)
    -devised tests with different and observable outcomes (train the rats to find food and then block the tunnel to the food to see if they can still find it)
    -collected evidence to exclude or include possible hypothesis
    -published
    -it is used in the field
    -performed with integrity
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Why do Science?

A

-I need to know something so that I can do something else (testing the unknown)

  • Other guy proved something. I need to verify this (e.g., we read text books and just trust everything they say but we should always want to verify what they’re saying is true even it goes along with other data)

= is the result “real”? CONFIRMATION

-need to learn a method

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

7 types of scientific questions

A
  1. Does something exist? (PRESENCE)
  2. what does something look like? (STRUCTURE)
  3. What is something made of? (COMPOSITION)
  4. What does something do? (FUNCTION)
  5. What makes something do something else? Is it because of something else completely different? (NETWORK)
  6. How does something do something else? (MECHANISM)
  7. wHAT DISRUPTS/CHANGES THE FUNCTION OF something? X
    How does another thing W –> Z?
    Does some W make X –> Y prime, not Y?

(MODULATION/ALTERATION)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Core properties investigated by 7 scientific questions

A

-description
-interaction
-how it works
-how it can be changed

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Mechanism research (q5 -7) vs Description study (q1-4)

A

-ascribing the wrong mechanism (blood move through the body by heat versus descriptive study of realizing heart pumps the blood through the body) is worse than no mechanism

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Testable items/theories

A
  1. Hypothesis - theory subjected to experimentation and falsification; doesn’t require info (should be falsifiable because it is deduction)
  2. Question - aimed to produce data and build (is open ended but must be closed to test and leads to hypothesis generation and building data/models - the overall goal)
  3. Model - a construct that uses all data generated to predict outcomes in future or different conditions because it is inductive
  4. Predictions - tool and ultimate goals (2 levels)

Difference between question and hypothesis = you can use a question to form a hypothesis and some research can’t accommodate hypothesis

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

MECHANISM Study

A

-understanding the how and the way something works

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Description Study

A

-characterizing something quantitatively or qualitatively

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Interaction Study

A

-communication between component of something to understand something else

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Modulation Study

A

-what happens when you change or alter something
-how something is regulated

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Deductive Inference

A

-often leads to syllogism

-(general observations used to make specific inferences/statements

  • deducing/conclusions comes from knowledge that is known or assumed
  • if the premise (e.g., background) is wrong then the deduction is wrong too)

-strip away incorrect facts until you’re left with truth

-inescapable logic

-topdown logic

  • example with aristotle/matematicians (he deduced that men have more teeth than women) - method = identifying differences between men than women and noticed that women are smaller than men in stature and so generally in animal kingdom it means they have fewer teeth)

-valid if impossible for premises to be true while conclusion is false (e.g., logic that makes sense but the data/premise is wrong - but might be valid even though the premises are untrue (you’re right but the situation is wrong)

-link the premises (theory, conditions) to the conclusions through a hypothesis

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Syllogism (categorical argument deduction)

A

-if A=B and B=C, then A=C
(attractive argument, popular)

-major premise: All books from the store are new
-minor premise: these books are from that store
-conclusion: therefore, these are new books

tricky example - All swans are white, a cygnet is grey, therefore cygnets are not swans (not true because cygnets are baby swans as it’s possible that swans are grey before they become white) = danger with deduction and hypothesis because what you observe doesn’t conform to what you believe is the case

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Inductive Inference

A
  • using past experiences to predict probability that an outcome will occur (future) if repeated
  • to infer a general law or principle from observation of particular instances (tests, trials, observations) - one inference that what is true in one case will be true in others that resemble it

-bottom up logic

-you take specific facts and you create the probability that something will happen (the point is to make you think that something is likely to happen but not that it’s a certainty)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

tacking a paper: main points

A
  • match the conclusions to the data (can you see the logic/predictions) – can you report on a paper merely based on the conclusions that they give? (no, because conclusions should match data as the data should support the conclusion

-is the data conducted in a way that is falsifiable?

-are there other possibilities (were there problems/flaws)

-are there technical issues (methodology?) (its probably done well since its published work and there are past peer review)

-if the conclusions are supported by the data, it’s probably a good paper

17
Q

Induction by Analogy

A

-observations of past experiences are generalized (predict results for something that is not identical to the original situation)

-e.g., clinical trial of a drug (3 out of 10 people respond = 30% efficacy therefore then using these results for saying you’d expect 30% of people to respond in the general population

18
Q

How Induction and Deduction go together

A

-induction gives you overall scientific goals and plan and then you create hypotheses and the hypotheses are something you can test with deductive reasoning

Deduction: Drawing a specific conclusion from general observation/premises (statement used to infer/propose a conclusion)

Induction: drawing general principles from specific observations/instances

-deduction’s conclusion is certain, even if the premise is wrong or weak

-inductive conclusions are probable - always involves uncertainty since you can’t see everything because you can’t know every outcome

  • inductive argument relies on a hypothesis that must be falsifiable

deduction: a proper (true) logical argument will be wrong if the premise is wrong

induction: while facts/observations may be true, there may not be an actual relationship between the observation and the conclusion (general) - you can’t know everything which is why this is probable and so you would falsify and strip away facts from parts or your entire working model

19
Q

Falsification

A

-when your logic is incorrect and it allows you to remove and exclude a piece of information

  • you only need one experiment to disprove a hypothesis
  • you never prove because if you fail to falsify, you don’t know for sure why it happened and so how do you have conclusive proof? experiements fail because humans are the weakest link
  • if thing (object, idea) +aspect (property) can be separated (e.g., you did 10 trials and you get the same effect but then you do an 11th trial and you can separate the effect), then you will either falsify at some point or, absence of evidence to reject implies it’s real = induction + probabilistic - the critic decides if it’s ok (critic is the person reading the paper)
  • it’s okay to predict future outcomes from past experiences as long as we have all the information
20
Q

Hypothetico-deductive model

A
  • Observations
  • then use induction (past observations, specific pieces of data) and make a hypothesis

-hypothesis sets up a line of reasoning that you can either falsify so you can exclude it or fail to falsify

-if you reject your hypothesis, then you use induction to create next round of hypotheses and repeat

  • if you fail to falsify, then you take the information and make another hypotheses since the first one wasn’t specific enough as now you can exclude something from it and keep excluding
21
Q

CAVEATS of the hypothetico-deductive method

A

-use induction to make a hypothesis that is falsifiable
-the hypothesis sets up a statement that is deductive (you can falsify or fail to falsify)
-predict the outcome if hypothesis is true (predictions help you to determine whether you’re seeing all possibilities)
-if it’s false, hypothesis is false
-if it’s not falsified, corroborates theory

22
Q

Why do you need 2 types of logic (deductive and inference)

A

-deductive inference is used to question data in an experiment (general observations -> specific conclusion, often using syllogism = prediction)

-inductive inference used to build a model and hypotheses and predictions (SPECIFIC observations brought together into a prediction of outcome (general, possibility)

  • togetherr = scientific method/caveats
23
Q

Questions as Hypothesis

A

-often used as hypothesis without “choosing sides”
-some research can’t accommodate hypotheses - use a question
-open ended arrangement but must be closed to test (e.g., what is the sequence of the human genome?)
-use when you have too little information, to screen for width of possibilities and to avoid bias or if you can’t make a statement of choice

24
Q

Types of Hypotheses

A
  • null hypothesis: no effect (statistical)
    -alternate hypothesis: different effect (has it’s own null)
    -negative hypothesis: opposite effect
  • because a hypothesis is a statement, it frequently becomes a conclusion
25
Q

Advantages to hypotheses in research design

A
  1. restrict your system to things that are testable and falsifiable
  2. just do the experiment
  3. you must be the skeptic
  4. data is within a larger package (from a larger study so there’s some data backing it) - the framework
  5. possible to maintain rigorous standard
26
Q

Hypotheses cause problems ***

A

1) demanding measurements on a limited or irrelevant subset of all data -> bias and misleading results

2) data becomes filtered

3) positive results become more valued than negatives

27
Q

Hallmarks of Models

A

-Come from data, should be specific
-use all the data (including failures)
-supported by evidence
-accuracy, convenience, consistency, repeatability allow it to predict the future
-only good if accurate, precise and real
- limited data -> limited power

28
Q

3 kinds of models

A

-simple: predict 1 situation, 1 condition (e.g., tree drops apples in fall, drug kills some lung cancer cells)

-compound: bring together variables, groups, often over time (tough if you have disagreement) - in snow/wind/rain, drug works in multiple trials and kills ltumours that are BRCA1

-generalized: model applied to different situations, times, settings, outcomes (MECHANISM, PERTURBATION) - HIGHEST STANDING AND TOUGH to attain - oranges?, drug will also work in prostate cancer?

29
Q

Two schools of thought (existential questions)

A
  • you can take data, and build a model and verify the model and modify it but are you looking at the right things

-you can take your model and make a hypothesis and then attempt to falsify it but hypothesis give bias since it’s a statement and you expect to see the outcome potentially and you also exclude some components that you falsify if they were actually important somewhere else

Molecular Biology (8 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions