mathvideo•This 180-second segment introduces a one-sample hypothesis-testing example about whether a drug affects rat response time. The typed problem gives n=100 injected rats, an untreated mean of 1.2 seconds, an observed sample mean of 1.05 seconds, and a sample standard deviation of 0.5 seconds. The instructor first formalizes the no-effect claim as H_0: = 1.2 s, then the competing effect claim as the two-sided alternative H_1: 1.2 s. The remainder of the clip explains the basic decision logic: assume H_0 is true, ask how probable the observed sample result would be under that assumption, and treat a very small probability as grounds to reject H_0. The segment stops before any numerical test statistic or explicit p-value computation is carried out. This 180-second whiteboard segment introduces a two-sided hypothesis test about whether a drug affects rat response time. The known no-drug mean is 1.2 s, while a sample of 100 injected rats has mean 1.05 s and sample standard deviation 0.5 s. The instructor assumes H_0 is true, draws the sampling distribution of the sample mean as approximately normal and centered at 1.2 s, derives the standard-error formula σ_{} = σ / √100, replaces the unknown σ with s, and computes = 0.05 s. The clip ends by identifying the next step as computing a z-score for 1.05 s relative to 1.2 s; the actual z-score and p-value are not completed within this excerpt. This 180-second whiteboard segment continues a hypothesis test about whether a drug affects rat response time. The board already states H0: μ = 1.2 s and H1: μ ≠ 1.2 s, with n = 100, observed x̄ = 1.05 s, s = 0.5 s, and the estimated standard deviation of the sample mean computed as 0.05 s. The instructor then forms a z-statistic, writes Z = (1.2 − 1.05)/0.05, simplifies it to 3, and interprets the observed sample mean as lying three standard deviations below the hypothesized mean on a bell-shaped sampling distribution. He marks ±1, ±2, and ±3 standard deviations, places 1.05 at the left −3 line, and explains that because the alternative is two-sided, “this extreme” includes both tails beyond ±3. Finally he invokes the empirical rule, labels the central region 99.7%, and sets up the complementary two-tail probability before the clip ends. This 146-second whiteboard segment completes a two-sided hypothesis test about whether a drug changes rat response time. The board already shows H_0: μ=1.2 s, H_1: μ≠1.2 s, a standard error of 0.05 s, and Z=3 from comparing the sample mean 1.05 s to the null mean. The speaker explains that the two tail areas together are 0.3%, or 0.003, and identifies this as the p-value: the probability of getting a result at least this extreme if the null hypothesis were true. Because 0.003 is far below the commonly mentioned 5% threshold, the lesson rejects H_0 and concludes that the drug has an effect, while also noting that statistical evidence is not 100% certainty.
This 180-second segment introduces a one-sample hypothesis-testing example about whether a drug affects rat response time. The typed problem gives n=100 injected rats, an untreated mean of 1.2 seconds, an observed sample mean of 1.05 seconds, and a sample standard deviation of 0.5 seconds. The instructor first formalizes the no-effect claim as H_0: = 1.2 s, then the competing effect claim as the two-sided alternative H_1: 1.2 s. The remainder of the clip explains the basic decision logic: assume H_0 is true, ask how probable the observed sample result would be under that assumption, and treat a very small probability as grounds to reject H_0. The segment stops before any numerical test statistic or explicit p-value computation is carried out. This 180-second whiteboard segment introduces a two-sided hypothesis test about whether a drug affects rat response time. The known no-drug mean is 1.2 s, while a sample of 100 injected rats has mean 1.05 s and sample standard deviation 0.5 s. The instructor assumes H_0 is true, draws the sampling distribution of the sample mean as approximately normal and centered at 1.2 s, derives the standard-error formula σ_{} = σ / √100, replaces the unknown σ with s, and computes = 0.05 s. The clip ends by identifying the next step as computing a z-score for 1.05 s relative to 1.2 s; the actual z-score and p-value are not completed within this excerpt. This 180-second whiteboard segment continues a hypothesis test about whether a drug affects rat response time. The board already states H0: μ = 1.2 s and H1: μ ≠ 1.2 s, with n = 100, observed x̄ = 1.05 s, s = 0.5 s, and the estimated standard deviation of the sample mean computed as 0.05 s. The instructor then forms a z-statistic, writes Z = (1.2 − 1.05)/0.05, simplifies it to 3, and interprets the observed sample mean as lying three standard deviations below the hypothesized mean on a bell-shaped sampling distribution. He marks ±1, ±2, and ±3 standard deviations, places 1.05 at the left −3 line, and explains that because the alternative is two-sided, “this extreme” includes both tails beyond ±3. Finally he invokes the empirical rule, labels the central region 99.7%, and sets up the complementary two-tail probability before the clip ends. This 146-second whiteboard segment completes a two-sided hypothesis test about whether a drug changes rat response time. The board already shows H_0: μ=1.2 s, H_1: μ≠1.2 s, a standard error of 0.05 s, and Z=3 from comparing the sample mean 1.05 s to the null mean. The speaker explains that the two tail areas together are 0.3%, or 0.003, and identifies this as the p-value: the probability of getting a result at least this extreme if the null hypothesis were true. Because 0.003 is far below the commonly mentioned 5% threshold, the lesson rejects H_0 and concludes that the drug has an effect, while also noting that statistical evidence is not 100% certainty.
Generated from the video's visuals and explanation; not verbatim speech.
The clip opens with a typed word problem on a black board. A neurologist injects 100 rats with a unit dose of a drug, applies a neurological stimulus, and records response times. The known mean response time for rats not injected with the drug is 1.2 seconds. For the 100 injected rats, the observed sample mean is 1.05 seconds and the sample standard deviation is 0.5 seconds. The question is whether the drug has an effect on response time.
To answer that question, the instructor says we need two hypotheses. The first is the null hypothesis, labeled H_0. He explains it as the status quo: assume the thing being studied has no effect. On the board this becomes the verbal statement 'Drug has no effect' and the symbolic statement H_0: = 1.2 s, where is the mean response time for rats taking the drug. The point is that if the drug truly does nothing, the drug-group mean should still match the untreated mean of 1.2 seconds.
Next he introduces the alternative hypothesis, labeled H_1. This is the claim that the drug does do something. He writes it as 'Drug has an effect' and then symbolically as H_1: 1.2 s when the drug is given. Because the inequality is two-sided, the alternative is not yet committed to the drug making responses only faster or only slower; it merely says the mean differs from 1.2 seconds.
With both hypotheses on the board, the instructor turns to the logic of the test. He asks how we should decide whether to accept the alternative or default back to the null because the data are not convincing enough. His answer is the standard hypothesis-testing strategy: temporarily assume H_0 is true. Then ask what probability there is of obtaining the observed sample result under that assumption. If that probability is very small, the result is surprising under the no-effect model, so we have reason to doubt H_0 and lean toward H_1. The clip ends just as he writes 'Assume H_0', setting up the next stage of the calculation.
The board presents a concrete hypothesis-testing problem: 100 rats receive a drug, their response times are recorded, and the known mean for uninjected rats is 1.2 seconds. The sample of injected rats has mean 1.05 seconds and sample standard deviation 0.5 seconds. The hypotheses are written explicitly as H_0: μ = 1.2 s and H_1: μ ≠ 1.2 s, so the test is two-sided and asks whether the drug changes the population mean response time.
The instructor then states the logic of the test: assume the null hypothesis is true, and ask how likely it would be to obtain the observed sample mean or something even more extreme. This is the key inferential move: the probability is computed under the no-effect model, not under the claim that the drug definitely works.
To make that probability question concrete, the lesson turns to the sampling distribution of the sample mean under H_0. A bell-shaped curve is drawn, and the speaker treats it as approximately normal because the sample size is 100. The center of this distribution is labeled μ_{} = μ = 1.2 s, reflecting the rule that the sampling distribution of the sample mean is centered at the population mean.
Next the spread of that sampling distribution is derived. The board writes σ_{} = σ / √100, which is the standard-error relationship for the sample mean. Because the population standard deviation σ is not known in this example, the instructor replaces it with the sample standard deviation s, yielding the approximation σ_{} ≈ s / √100.
The numerical substitution is then carried out step by step. With s = 0.5 and √100 = 10, the estimated standard error is written as = 0.5 / 10 = 0.05 seconds. The hat notation is explained as marking an estimate formed by using the sample standard deviation in place of the unknown population standard deviation.
At this point the lesson has both ingredients needed to standardize the observed mean: the null-centered mean 1.2 s and the estimated standard error 0.05 s. The instructor asks how many standard deviations away 1.05 s lies from 1.2 s and identifies that question as computing a z-score. The clip stops at this setup stage, so the actual z-value and the subsequent p-value calculation are not shown within the provided 180 seconds.
The board presents a concrete hypothesis-testing scenario: 100 rats are injected with a drug, the known mean response time without the drug is 1.2 seconds, and the injected sample has mean 1.05 seconds with sample standard deviation 0.5 seconds. The hypotheses are already written as H0: μ = 1.2 s and H1: μ ≠ 1.2 s, so the question is whether the drug changes response time in either direction.
A bell curve labeled with center μx̄ = μ = 1.2 s represents the sampling distribution assumed under H0. To the right, the standard deviation of the sample mean has already been estimated as σ̂x̄ = σ̂/√100 ≈ 0.5/√100 = 0.05. This prepares the scale needed to judge how unusual the observed sample mean is.
The instructor now computes a standardized score. He writes Z = (1.2 − 1.05)/0.05, explaining that the numerator measures the distance from the hypothesized mean and the denominator converts that distance into standard-deviation units. He deliberately orders the subtraction as 1.2 − 1.05 so the displayed distance is positive.
The arithmetic is simplified on screen to Z = 0.15/0.05 = 3. Thus the observed sample mean is three estimated standard deviations away from the hypothesized mean in magnitude. Because 1.05 is less than 1.2, the corresponding location on the curve is on the left side, below the mean.
To make that location visible, the instructor draws dashed vertical markers at one, two, and three standard deviations on both sides of the center. The observed value 1.05 is then placed at the third negative standard-deviation line, visually confirming that the sample result is three standard deviations below the hypothesized mean.
Next he asks for the probability of getting a result this extreme by chance. Since the alternative hypothesis is two-sided, “this extreme” is not only values below 1.05; it also includes values at least three standard deviations above the mean. Accordingly, both far tails beyond ±3 standard deviations are shaded.
Using the empirical rule, he states that about 99.7% of the probability lies within three standard deviations of the mean, and labels the central region 99.7%. The remaining probability, concentrated in the two shaded tails, is therefore found by complement: 1 − 0.997 = 0.003, or about 0.3% total in the two tails.
The clip opens on a completed whiteboard setup for a hypothesis test about a drug's effect on rat response time. The printed problem states that untreated rats have mean response time 1.2 seconds, while 100 injected rats have sample mean 1.05 seconds and sample standard deviation 0.5 seconds. The handwritten hypotheses are H_0: μ=1.2 s and H_1: μ≠1.2 s, so the test is two-sided.
On the right, the board computes the standard error of the sample mean as σ_{x̄}≈0.5/√100=0.05. On the left, under 'Assume H_0', it standardizes the observed mean: Z=(1.2-1.05)/0.05=0.15/0.05=3. The bell curve in the middle is centered at μ_{x̄}=μ=1.2 s, with the large central region labeled 99.7% and the two tails marked as the remaining extreme region.
The speaker now focuses on those tails. He says that if the middle accounts for 99.7%, then the two tails combined account for 0.3%, and he rewrites that as the decimal 0.003. Visually, arrows and labels emphasize that both tails together represent the probability of getting a result at least as extreme as the observed one when H_0 is true.
He interprets 0.003 as 'less than 1 in 300'. The logic is: assuming the drug has no effect, a sample mean this far from 1.2 seconds would happen with probability only 0.003. That makes the observed result very surprising under the null hypothesis, so the evidence points away from H_0 and toward H_1.
The board is annotated with the word 'reject' near H_0 as the speaker states his decision. He explicitly says he is going to reject the null hypothesis, while also acknowledging that this is not 100% certainty. The conclusion is probabilistic: the data are too unlikely under H_0, so the alternative hypothesis is favored.
Next, the lesson introduces terminology. The speaker says that this tail probability, the probability of getting a result more extreme than the observed one given the null hypothesis, is called a p-value. He writes 'P-value :' at the bottom and fills in '.003', linking the named concept directly to the earlier curve-area calculation.
With p=0.003 established, he compares it to a common decision threshold. He says that many people use 5%, meaning a 1-in-20 chance, as the cutoff for rejecting H_0. Since 0.003 is much smaller than 0.05, this example falls well past that threshold and provides strong evidence against the no-effect hypothesis.
The final takeaway is that the drug appears to affect response time. The speaker circles H_1 and summarizes that the null hypothesis is rejected because the observed sample mean is 3 standard errors away from the hypothesized mean, corresponding to a p-value of 0.003. The segment ends by reinforcing both the computational result and the interpretation that a very small p-value supports the alternative hypothesis.
A researcher injects 100 rats with a drug and records response times. The untreated mean is known to be 1.2 seconds. The injected sample has mean 1.05 seconds and sample standard deviation 0.5 seconds. The statistical question is whether the drug changes the mean response time.
The null hypothesis is the no-effect baseline. In this example it says the drug does not change response time, so the population mean for treated rats is still 1.2 seconds.
The alternative hypothesis is the competing claim that the drug has some effect. Here it is written as a two-sided inequality, meaning the treated mean differs from 1.2 seconds in either direction.
After stating H_0 and H_1, the method is to assume H_0 is true and ask how likely the observed sample result would be under that assumption. If the observed result is very unlikely under H_0, that gives grounds to reject H_0 and favor H_1. This clip sets up that logic but does not yet perform the numerical probability calculation.
The example tests whether a drug changes rat response time. The null hypothesis says there is no effect, so the population mean stays at the known uninjected value μ = 1.2 s. The alternative says the drug has an effect, so μ ≠ 1.2 s. The use of ≠ makes this a two-sided test.
From the 100 injected rats, the observed sample mean is 1.05 seconds and the observed sample standard deviation is 0.5 seconds. These are the data that will be judged against the null-hypothesis model.
The instructor explicitly conditions the analysis on the null hypothesis being true. The question is not merely the chance of getting exactly 1.05 s, but the chance of getting a result like 1.05 s or more extreme under the no-effect model.
Under H_0, the sampling distribution of the sample mean is drawn as approximately normal and centered at the null population mean. The board labels this center as μ_{} = μ = 1.2 s.
The spread of the sampling distribution of the sample mean is the population standard deviation divided by the square root of the sample size. For this example, that is written as σ_{} = σ / √100.
Because the population standard deviation is not known, the lesson substitutes the sample standard deviation s for σ. The speaker justifies this as reasonable with a large sample size, here n = 100.
Plugging in s = 0.5 and √100 = 10 gives an estimated standard error of 0.05 seconds. The hat notation marks this as an estimate rather than the exact population-based quantity.
Once the center and standard error are known, the instructor reframes the problem as asking how many standard deviations 1.05 s is from 1.2 s. That standardized distance is a z-score. The clip sets up this step but does not finish the arithmetic.
The test compares a known no-drug mean response time of 1.2 seconds with a sample of 100 injected rats whose mean is 1.05 seconds. The null hypothesis is H0: μ = 1.2 s, meaning the drug has no effect. The alternative is H1: μ ≠ 1.2 s, meaning the drug changes the mean in either direction.
Before standardizing the result, the video estimates the spread of the sampling distribution of the mean using the sample standard deviation 0.5 and sample size 100. This gives 0.05 seconds, which becomes the denominator of the z-statistic.
The z-statistic measures how many estimated standard deviations the observed sample mean lies from the hypothesized mean. Here the instructor writes the distance as 1.2 − 1.05 and divides by 0.05 to standardize the discrepancy.
Substituting the numbers yields Z = (1.2 − 1.05)/0.05 = 0.15/0.05 = 3. The magnitude 3 means the sample mean is three standard deviations from the hypothesized mean; because 1.05 < 1.2, it is on the lower side.
The bell curve is centered at 1.2. After marking ±1, ±2, and ±3 standard deviations, the observed value 1.05 is placed at the third negative standard-deviation line, showing visually that the sample result is three standard deviations below the mean.
Because H1 is μ ≠ 1.2, an outcome this extreme includes both unusually low sample means and unusually high sample means. The video therefore shades both tails beyond three standard deviations rather than only the left tail where the observed value lies.
The empirical rule says about 99.7% of a normal distribution lies within three standard deviations of the mean. The central region is labeled 99.7%, so the combined probability in the two outer tails is 1 − 0.997 = 0.003, or about 0.3%.
The example tests whether a drug changes mean response time. The null hypothesis states no change, μ=1.2 seconds. The alternative states a change in either direction, μ≠1.2 seconds, which is why both tails of the distribution matter.
Using the displayed spread 0.5 and sample size 100, the board computes the standard deviation of the sampling distribution of the sample mean as 0.5/√100=0.05 seconds. This becomes the scale used to judge how unusual the observed sample mean is.
The observed sample mean 1.05 is compared to the null mean 1.2 by subtracting and dividing by the standard error 0.05. The arithmetic shown is (1.2-1.05)/0.05=0.15/0.05=3, so the sample mean is 3 standard errors below the null center.
The normal-curve sketch labels the central region as 99.7%. The speaker takes the complement to conclude that the two tails together have probability 0.3%, which he rewrites as 0.003 in decimal form.
The p-value is defined verbally as the probability, assuming the null hypothesis is true, of getting a result at least as extreme as the observed one. Here that probability is exactly the combined tail area 0.003.
Explore conditions, steps and evidence. Supplementary explanations are labeled separately from content shown in the video.
Handwritten yellow label H_0 is written on the board and remains visible.
The speaker calls it the null hypothesis.
H_0
Null hypothesis for this example: the drug has no effect on response time.
Statistical hypothesis label.
Handwritten green label H_1 is written below H_0 and remains visible.
The speaker introduces it as the alternative hypothesis.
H_1
Alternative hypothesis for this example: the drug has an effect on response time.
Statistical hypothesis label.
Greek letter μ appears in both handwritten hypothesis statements.
The speaker refers to it as the mean response time with the drug.
Population mean response time for rats given the drug, measured in seconds.
Real-valued parameter representing a population mean.
Typed problem text states 'injecting 100 rats'.
The speaker reads that 100 rats were injected.
n
Sample size, here the number of injected rats.
Positive integer; in this example n=100.
Typed problem text gives 'The mean of the 100 injected rats' response times is 1.05 seconds'.
The speaker reads the sample mean as 1.05 seconds.
Sample mean response time of the 100 injected rats.
Real number; in this example =1.05 seconds.
Typed problem text gives 'with a sample standard deviation of 0.5 seconds'.
The speaker reads the sample standard deviation as 0.5 seconds.
s
Sample standard deviation of the injected rats' response times.
Positive real number; in this example s=0.5 seconds.
Typed problem text states the mean response time for rats not injected with the drug is 1.2 seconds.
The handwritten hypotheses use 1.2 s as the reference value.
Reference population mean response time without the drug, used as the hypothesized value under H_0.
Real number; in this example =1.2 seconds.
Handwritten yellow text begins 'Assume H_0' near the end of the clip.
The speaker says, 'Let's assume that the null hypothesis is true.'
Only the beginning of the sentence is visible before the clip ends.
Assume H_0
Opening statement of the next reasoning step: temporarily take the null hypothesis as true in order to evaluate how surprising the observed sample result would be.
Methodological assumption in hypothesis testing.
Written on screen as H_0: Drug has no effect ⇒ μ = 1.2 s (even w/ drug).
The speaker repeatedly says to assume the null hypothesis is true.
H_0
Null hypothesis for this test: the drug has no effect on response time.
Statement about the population mean response time under the no-effect model.
Written on screen as H_1: Drug has an effect ⇒ μ ≠ 1.2 s when the drug is given.
The speaker frames the test as checking whether the drug has an effect.
H_1
Alternative hypothesis: the drug changes the population mean response time.
Two-sided statement about the population mean response time when the drug is administered.
Appears in H_0 and H_1 as μ = 1.2 s and μ ≠ 1.2 s.
The speaker refers to the mean of the population distribution equal to 1.2 seconds.
μ
Population mean response time for rats receiving the drug.
Measured in seconds.
The problem text states the mean of the 100 injected rats' response times is 1.05 seconds.
The speaker asks for the probability of getting a sample mean of 1.05 seconds.
Sample mean response time from the 100 injected rats.
Observed value in this example is 1.05 seconds.
Typed problem text describes a neurologist injecting 100 rats with a unit dose of a drug, recording response times, comparing them with a known no-drug mean of 1.2 seconds, and giving sample mean 1.05 seconds and sample standard deviation 0.5 seconds.
The speaker reads the same problem aloud and asks whether the drug has an effect on response time.
The clip sets up a one-sample mean comparison problem. A researcher injects 100 rats, records their response times, and wants to decide whether the drug changes the mean response time relative to the known untreated mean of 1.2 seconds. The observed sample statistics are a mean of 1.05 seconds and a sample standard deviation of 0.5 seconds.
The question is whether the drug has any effect on response time.
The untreated mean response time is taken as known.
The sample consists of 100 injected rats.
Yellow handwriting writes H_0: Drug has no effect => = 1.2 s (even w/ drug).
The speaker says the first hypothesis is the null hypothesis, that the drug has no effect, and explains it as the status quo assumption.
The null hypothesis is the default claim that the treatment does nothing. In this example it is stated verbally as 'Drug has no effect' and symbolically as the population mean response time with the drug still being 1.2 seconds, the same as without the drug.
Applies to the specific example about drug effect on rat response time.
The parameter denotes the mean response time for rats given the drug.
The value 1.2 s is the known untreated mean used as the reference.
Green handwriting writes H_1: Drug has an effect => 1.2 s when the drug is given.
The speaker introduces the alternative hypothesis as the claim that the drug actually does do something.
The alternative hypothesis is the competing claim that the treatment does have an effect. Here it is two-sided: the mean response time with the drug is not equal to 1.2 seconds. The clip does not specify a direction such as faster or slower only; it simply says the drug has an effect.
Paired with H_0 in the same example.
The inequality is two-sided, not one-sided.
The parameter again denotes the mean response time for rats given the drug.
The speaker asks how we know whether to accept the alternative or default to the null because the data is not convincing.
He then says the method is to assume the null hypothesis is true and ask what probability there is of getting these sample results; if that probability is really small, then the null probably isn't true and we can reject it.
Near the end he writes 'Assume H_0'.
The clip introduces the logic but does not yet compute the probability or define the term p-value explicitly within this 180-second segment.
The video presents the core reasoning pattern of hypothesis testing: temporarily assume H_0 is true, then ask how likely the observed sample outcome would be under that assumption. If the observed result would be very unlikely under H_0, that provides grounds to doubt H_0 and favor H_1. This segment stops at the setup of that logic rather than carrying out the calculation.
The test is based on evaluating the sample result under the assumption that H_0 is true.
A very small probability of the observed result under H_0 motivates rejection of H_0.
This clip gives the conceptual method, not the completed numerical test.
Screen shows H_0: Drug has no effect ⇒ μ = 1.2 s (even w/ drug) and H_1: Drug has an effect ⇒ μ ≠ 1.2 s when the drug is given.
The speaker begins by saying, 'And so if we assume the null hypothesis is true...'
This segment sets up a two-sided hypothesis test about the drug's effect on response time. The null hypothesis states that the drug has no effect, so the population mean remains μ = 1.2 s even when the drug is given. The alternative hypothesis states that the drug has an effect, meaning the population mean is not equal to 1.2 s when the drug is given.
The parameter being tested is the population mean response time μ.
The comparison value 1.2 s is the known mean for rats not injected with the drug.
The alternative is two-sided because it uses ≠.
Problem text at top states 100 rats, mean response time for rats not injected is 1.2 seconds, mean of the 100 injected rats is 1.05 seconds, and sample standard deviation is 0.5 seconds.
The speaker repeats the observed sample mean 1.05 seconds and standard deviation 0.5 seconds.
The example compares a known baseline mean response time of 1.2 seconds for uninjected rats with a sample of 100 injected rats whose observed mean response time is 1.05 seconds and whose sample standard deviation is 0.5 seconds. The question is whether the drug has an effect on response time.
Baseline population mean under H_0 is 1.2 s.
Sample size is 100 injected rats.
Observed sample mean is 1.05 s.
Observed sample standard deviation is 0.5 s.
The speaker says, 'let's just think about the sampling distribution if we assume the null hypothesis.'
A bell-shaped curve is drawn with a vertical dashed center line.
Below the curve the screen writes μ_{} = μ = 1.2 s.
Under the assumed null hypothesis, the sampling distribution of the sample mean is centered at the population mean specified by H_0. In this example that center is 1.2 seconds, so the mean of the sampling distribution is μ_{} = μ = 1.2 s.
Assume H_0 is true.
The distribution being discussed is the sampling distribution of the sample mean.
The population mean under H_0 is 1.2 s.
The speaker says the standard deviation of the sampling distribution should be equal to the standard deviation of the population distribution divided by the square root of the sample size.
Screen writes σ_{} = σ / √100.
The spread of the sampling distribution of the sample mean is the population standard deviation divided by the square root of the sample size. Here the sample size is 100, so the formula is written as σ_{} = σ / √100.
Applies to the sampling distribution of the sample mean.
n is the sample size.
σ is the population standard deviation of individual observations.
The speaker says, 'We do not know what the standard deviation of the entire population is. So what we're going to do is estimate it with our sample standard deviation.'
Screen rewrites the expression as approximately s / √100.
Because the population standard deviation σ is unknown, the lesson replaces it with the sample standard deviation s. The speaker explicitly says this is reasonable because the sample size is large, greater than 100, making s a good approximator for σ in this context.
σ is unknown.
s is available from the sample.
The speaker justifies the approximation by the large sample size.
Screen shows = 0.5 / √100 = 0.5 / 10 = 0.05.
The speaker computes 0.5 divided by 10 and says the result is 0.05.
Using s = 0.5 and n = 100, the estimated standard deviation of the sampling distribution is computed as 0.5 divided by the square root of 100. Since √100 = 10, the value is 0.5 / 10 = 0.05 seconds. The hat notation indicates that this is an estimate rather than the exact population-based quantity.
s = 0.5 s.
n = 100.
The result is expressed in seconds.
The speaker asks how many standard deviations away from the mean 1.05 seconds is, and then says, 'essentially we're just figuring out a z-score.'
The bell curve remains on screen with center labeled μ_{} = μ = 1.2 s and the estimated standard error 0.05 written nearby.
The actual z-score calculation is not completed within this clip.
After establishing the center and spread of the sampling distribution, the lesson reframes the question as asking how many standard deviations the observed sample mean 1.05 s lies from the hypothesized center 1.2 s. The speaker identifies this as computing a z-score, which is the next step toward assessing the probability of a result at least that extreme.
Use the observed sample mean =1.05 s.
Use the null-hypothesis center μ_{}=1.2 s.
Use the estimated standard error =0.05 s.
The clip states the method but does not finish the arithmetic.
H_0: Drug has no effect => μ = 1.2 s (even w/ drug).
H_1: Drug has an effect => μ ≠ 1.2 s when the drug is given.
The setup frames the question as whether the drug has an effect on response time.
The video defines the null hypothesis as no drug effect, meaning the population mean response time remains 1.2 seconds even with the drug. The alternative hypothesis is two-sided: the drug has an effect, so the mean is not equal to 1.2 seconds when the drug is given.
Testing whether a drug affects response time.
Alternative is two-sided because the question asks whether there is any effect, not only an increase or decrease.
The speaker says the null hypothesis is always going to be viewed as the status quo, assuming whatever you're researching has no effect.
This is presented as explanatory intuition rather than a formal theorem.
In this lesson, the null hypothesis is interpreted as the status quo assumption that the treatment being studied has no effect.
The context is a basic introduction to hypothesis testing.
The example concerns whether a drug affects response time.
Described generally by the speaker for the null hypothesis in this introductory setting.
The speaker says that if the null hypothesis was true, we ask the probability of getting these results with the sample; if that probability is really, really small, then the null hypothesis probably isn't true and we could reject it.
The clip does not specify a significance threshold or formally define the probability quantity being discussed.
If, assuming H_0 is true, the observed sample result has a very small probability, then the lesson treats that as grounds to reject H_0 and lean toward H_1.
H_0 is assumed true for the purpose of evaluating the sample outcome.
The observed sample result is fixed from the experiment.
Informal 'really, really small' probability criterion; no exact cutoff is stated in this clip.
The speaker says, 'if we assume the null hypothesis is true, let's try to figure out the probability that we would have actually gotten this result... or even more extreme than this.'
To evaluate the evidence against the null hypothesis, the lesson first assumes H_0 is true and then asks for the probability of obtaining the observed sample mean or something more extreme.
Assume H_0: μ = 1.2 s.
Observed sample mean is 1.05 s from n = 100.
The test is framed around results at least as extreme as the observed one.
For the given sample result, compute the probability under the assumption that H_0 holds.
The speaker says, 'It'll be a normal distribution. We have a good number of samples, we have 100 samples here, so this is the sampling distribution.'
A symmetric bell-shaped curve is drawn to represent the sampling distribution.
The clip does not state the formal theorem name or all conditions beyond having 100 samples.
With 100 samples, the speaker treats the sampling distribution of the sample mean as a normal distribution.
Sample size is 100.
The quantity considered is the sampling distribution of the sample mean.
In this example, because n = 100 is described as a good number of samples, the sampling distribution is modeled as normal.
The speaker says, 'we'll put a little hat over it to show that we approximated the population standard deviation with the sample standard deviation.'
Screen writes = 0.05.
The hat on σ_{} indicates that the displayed value is an estimate obtained by substituting the sample standard deviation for the unknown population standard deviation.
σ is unknown.
s is used in its place.
The resulting quantity is written with a hat.
For this example, denotes the estimated standard deviation of the sampling distribution.
Z = .15/.05 = 3.
Speaker says this is going to be 3.
For this sample, the standardized distance from the hypothesized mean is 3 estimated standard deviations.
Hypothesized mean μ = 1.2 s.
Observed sample mean x̄ = 1.05 s.
Estimated standard deviation of the sample mean = 0.05 s.
For the given sample of 100 injected rats.
This result right here is three standard deviations away from the mean... three standard deviations below the mean.
The sample result is marked on the left side of the bell curve at the third negative standard-deviation line.
The observed value 1.05 seconds is located three estimated standard deviations below the hypothesized mean 1.2 seconds on the sampling-distribution curve.
Z = 3 in magnitude.
Observed mean is less than hypothesized mean.
For this specific sample result.
What is the probability of getting a result this extreme by chance?... it could be either a result less than this or a result that extreme in the positive direction, more than three standard deviations.
Both far tails beyond ±3 standard deviations are shaded magenta.
Because the alternative hypothesis is two-sided, a result this extreme includes values at least three standard deviations below the mean or at least three standard deviations above the mean.
H_1 is μ ≠ 1.2 s.
Observed |Z| = 3.
Over repeated samples under H_0.
Well these are 99.7% then both of these combined are going to be 0.3%.
Curve labels show 99.7% in the middle and 0.3% / .003 in the tails.
If the central region under the curve accounts for 99.7% of the total area, then the two remaining tails together account for 0.3%, i.e. 0.003 as a decimal.
The curve is normalized so total area is 1.
The displayed 99.7% refers to the central non-tail region.
For the specific normal-curve diagram shown in this example.
So at least from my point of view, this result seems to favor the alternative hypothesis.
I am going to reject the null hypothesis.
...the drug definitely has some effect.
The speaker frames the conclusion partly as judgment ('at least from my point of view') before stating rejection.
Given the observed sample mean 1.05, standard error 0.05, resulting Z=3, and p-value 0.003, the lesson concludes that the null hypothesis should be rejected and the alternative hypothesis favored.
H_0: μ=1.2.
H_1: μ≠1.2.
p-value is computed as 0.003.
Rejection threshold discussed is 5%.
For this rat-response-time example.
The speaker says they will set up two hypotheses.
He writes H_0 in words and symbols, then writes H_1 in words and symbols.
Start from the typed problem statement asking whether the drug has an effect on response time.
Given directly by the problem text and narration.
Formulate the default claim that the drug changes nothing.
Stated by the speaker as the null hypothesis.
Translate the verbal null hypothesis into a statement about the population mean response time with the drug.
The speaker says the mean with the drug should still be 1.2 seconds even with the drug.
Formulate the competing claim that the drug does do something.
Stated by the speaker as the alternative hypothesis.
Translate the verbal alternative hypothesis into a two-sided inequality about the same population mean.
The speaker writes and says the mean does not equal 1.2 seconds when the drug is given.
The example is converted from a plain-language research question into the paired statistical hypotheses H_0: = 1.2 s and H_1: 1.2 s.
The speaker explains that to decide between H_0 and H_1, assume H_0 is true and ask how probable the observed sample result would be under that assumption.
He writes 'Assume H_0' at the end of the clip.
The actual probability calculation is not performed within this segment.
After writing both hypotheses, the speaker turns to the question of how to decide which one the data support.
Explicitly asked in the narration.
Temporarily take the no-effect hypothesis as the working model.
The speaker says this is the way the test will be done and writes 'Assume H_0'.
Ask what probability there is of getting these sample results if H_0 were true.
Stated verbally as the next evaluative question.
Use the smallness of that conditional probability as evidence against the null and in favor of the alternative.
The speaker says that if the probability is really small, then the null probably isn't true and we could reject it.
The clip establishes the conceptual route to a p-value-style test: assume H_0, evaluate how surprising the sample is under that assumption, and reject H_0 when that surprise is sufficiently large.
Screen successively shows σ_{} = σ / √100, then ≈ s / √100, then = 0.5 / √100 = 0.5 / 10 = 0.05.
The speaker explains each substitution and computes the final decimal.
Start from the general relation between the standard deviation of the sampling distribution and the population standard deviation.
Stated directly in the video as the standard deviation of the sampling distribution equals the population standard deviation divided by the square root of the sample size.
Replace the unknown σ with the sample standard deviation s.
The speaker says σ is unknown and estimates it with s, calling this reasonable because the sample size is large.
Insert the numerical sample standard deviation s = 0.5 and mark the quantity as an estimate with a hat.
The problem gives s = 0.5 s, and the speaker explicitly adds a hat to indicate approximation.
Evaluate the square root and divide.
Arithmetic shown on screen and spoken aloud in the video.
The estimated standard deviation of the sampling distribution is = 0.05 seconds.
The speaker asks how many standard deviations away from the mean 1.05 seconds is and says this is essentially figuring out a z-score.
The needed quantities are already on screen: =1.05, μ_{}=1.2, and =0.05.
The clip does not write the full z-score formula or compute the final numeric z-value.
Collect the observed sample mean, the null-hypothesis center, and the estimated standard error.
These values are explicitly present on screen by this point in the clip.
Express the distance from the mean in standard-deviation units.
The speaker identifies the next task as finding how many standard deviations away 1.05 is, i.e. a z-score; the formula is the standard definition corresponding to that description.
The clip sets up the next step as computing a z-score from the observed mean, the null mean, and the estimated standard error, but does not complete the calculation within the provided duration.
Z = (1.2 - 1.05)/0.05.
Z = .15/.05 = 3.
Speaker explains subtracting to get a positive distance and dividing by the sampling-distribution standard deviation.
Identify the hypothesized mean, observed sample mean, and estimated standard deviation of the sample mean from the board.
These values are written in the problem setup and prior calculations.
Form the standardized distance by subtracting the observed mean from the hypothesized mean and dividing by the estimated standard deviation.
Definition of the z-statistic used in the video.
Compute the numerator 1.2 − 1.05 = 0.15.
Arithmetic simplification.
Divide 0.15 by 0.05 to obtain 3.
Arithmetic simplification.
The sample mean is 3 estimated standard deviations away from the hypothesized mean in magnitude.
We know from the empirical rule that 99.7% of the probability is within three standard deviations.
Central region labeled 99.7%; both tails shaded.
The clip stops before the speaker explicitly states the final numeric tail total.
Within three standard deviations of the mean lies about 99.7% of the normal probability.
Empirical rule stated in the audio.
The probability outside three standard deviations is the complement of the central probability.
Total probability under the curve is 1.
Subtract 0.997 from 1 to get the combined two-tail probability.
Arithmetic complement calculation.
The two shaded tails together represent approximately 0.3% of the probability under the sampling distribution.
or pink areas. Well these are 99.7% then both of these combined are going to be 0.3%.
Central shaded region labeled 99.7%; two tails labeled 0.3% and .003.
The video does not recompute the 99.7% from first principles inside this clip; it uses the already drawn curve values.
The diagram marks the large central portion of the bell curve as 99.7%.
Directly read from the board.
The two pink tails together are the complement of the central region.
Total area under the probability curve is 1, so tail area is 1 minus central area.
The speaker rewrites the percentage as a decimal probability.
Percent-to-decimal conversion.
The probability associated with results at least as extreme as the observed one is 0.003.
σ_{x̄} ≈ σ/√100 = 0.5/√100 = 0.5/10 = 0.05.
Z = (1.2 - 1.05)/0.05 = .15/.05 = 3.
The clip does not explicitly discuss why a Z-based normal approximation is appropriate beyond the displayed setup.
Insert the given spread 0.5 and sample size 100 into the standard-error formula.
Displayed formula on the right side of the board.
Evaluate √100=10 and divide.
Arithmetic shown on the board.
Subtract the observed sample mean from the null-hypothesis mean and divide by the standard error.
Formula written under 'Assume H_0'.
Simplify numerator and quotient to obtain the standardized distance.
Arithmetic shown on the board.
The observed sample mean lies 3 standard errors below the null-hypothesis mean.
If we assume that the drug has no effect, the probability of getting a sample this extreme or actually more extreme than this is only 0.3% ... less than 1 in 300.
If you have a p-value less than 5% ... I'm going to reject the null hypothesis.
P-value : .003.
The phrase 'definitely has some effect' is stronger than a formal statistical guarantee; the clip itself also says 'I don't know 100% sure'.
The tail probability from the curve is identified as the p-value.
Audio definition plus board label P-value : .003.
The p-value is much smaller than the commonly mentioned 5% cutoff.
Numerical comparison of 0.003 and 0.05.
Because the result is very unlikely under H_0, the lesson chooses the alternative hypothesis.
Spoken decision rule and explicit statement 'I am going to reject the null hypothesis.'
The example rejects the no-effect hypothesis and concludes that the drug has some effect on response time.
Typed problem text gives the full scenario with 100 rats, untreated mean 1.2 seconds, injected sample mean 1.05 seconds, and sample standard deviation 0.5 seconds.
The board later shows H_0: = 1.2 s and H_1: 1.2 s.
The speaker reads the scenario and uses it to introduce hypothesis testing.
No final numerical test statistic or probability is computed within this clip.
A neurologist injects 100 rats with a unit dose of a drug, records their response times, and wants to know whether the drug has an effect on response time given that untreated rats have mean response time 1.2 seconds.
n=100 injected rats
Known untreated mean response time = 1.2 seconds
Observed sample mean of injected rats = 1.05 seconds
Observed sample standard deviation = 0.5 seconds
Set up the statistical question of whether the drug affects response time and explain the logic for deciding between no effect and some effect.
State the no-effect hypothesis as equality of the drug-group population mean to the untreated mean.
Directly written on the board and explained verbally.
State the two-sided effect hypothesis as inequality from 1.2 seconds.
Directly written on the board and explained verbally.
Begin the testing logic by treating the null as true and considering how likely the observed sample mean would be under that assumption.
The speaker says to assume the null is true and asks for the probability of getting these results with the sample; the explicit conditioning notation is a faithful formalization of that spoken idea.
The clip does not reach a numerical conclusion. It concludes by formulating H_0 and H_1 and starting the method of assuming H_0 to judge how surprising the sample result is.
The written hypotheses match the spoken explanation, and the typed problem values 100, 1.2, 1.05, and 0.5 remain consistent throughout the segment.
Top text gives the full scenario: 100 rats, baseline mean 1.2 s, sample mean 1.05 s, sample standard deviation 0.5 s, and the question whether the drug has an effect.
The speaker works through the setup under the assumption that the null hypothesis is true.
The final p-value and decision are not reached within this clip.
A neurologist injects 100 rats with a drug and records response times. The known mean response time for rats not injected is 1.2 seconds. The injected sample has mean 1.05 seconds and sample standard deviation 0.5 seconds. Does the drug have an effect on response time?
n = 100
Baseline mean under H_0: μ = 1.2 s
Observed sample mean: = 1.05 s
Observed sample standard deviation: s = 0.5 s
Hypotheses: H_0: μ = 1.2 s; H_1: μ ≠ 1.2 s
Begin the hypothesis test by assuming H_0 is true, modeling the sampling distribution of the sample mean, estimating its standard error, and preparing to measure how extreme the observed mean is.
State the null and alternative hypotheses.
Written directly on the board at the start of the clip.
Condition the probability calculation on the null hypothesis being true.
Spoken explicitly by the instructor.
Center the sampling distribution of the sample mean at the null-hypothesis population mean.
Stated in audio and written below the bell curve.
Write the standard error formula and substitute the sample standard deviation for the unknown population standard deviation.
Explained verbally and shown symbolically on screen.
Compute the estimated standard error numerically.
Shown as a sequence of algebraic substitutions on the board and read aloud.
Set up the next step as measuring how many standard deviations 1.05 s is from 1.2 s.
The speaker says this is essentially figuring out a z-score; the formula is the standard expression matching that description.
By the end of the clip, the setup yields a null-centered normal sampling distribution with mean 1.2 s and estimated standard error 0.05 s, and the instructor identifies the next step as computing a z-score for the observed mean 1.05 s.
The numerical values match the on-screen problem statement and the written computations: 0.5 / √100 = 0.5 / 10 = 0.05, and the center is labeled μ_{} = μ = 1.2 s.
Problem text: 100 rats, baseline mean 1.2 seconds, sample mean 1.05 seconds, sample standard deviation 0.5 seconds.
Speaker works through the z-statistic and interprets the result on the bell curve.
A neurologist injects 100 rats with a unit dose of a drug and records response times. Rats not injected have mean response time 1.2 seconds. The 100 injected rats have mean 1.05 seconds and sample standard deviation 0.5 seconds. Does the drug have an effect on response time?
n = 100
Baseline/hypothesized mean μ = 1.2 s
Observed sample mean x̄ = 1.05 s
Sample standard deviation s = 0.5 s
H_0: μ = 1.2 s
H_1: μ ≠ 1.2 s
Standardize the observed sample mean and interpret how extreme it is under H_0.
Estimate the standard deviation of the sampling distribution of the mean.
Formula shown on the board for the standard error estimate.
Compute the z-statistic as the distance from the hypothesized mean divided by the estimated standard deviation.
Definition of z-statistic used in the clip.
Place the observed sample mean on the left tail of the bell curve at −3 standard deviations.
Audio interpretation plus diagram placement.
Use the empirical rule to identify the central probability within three standard deviations.
Explicitly stated in the audio.
Take the complement to find the combined probability in both tails beyond three standard deviations.
Derived from total probability 1 and the empirical-rule value.
The observed sample mean is 3 estimated standard deviations below the hypothesized mean, and the two-tail extreme region corresponds to about 0.3% probability under H_0.
Check that 1.2 − 1.05 = 0.15 and 0.15 / 0.05 = 3; verify that the central 99.7% leaves 0.3% in the two tails.
Problem text: A neurologist is testing the effect of a drug on response time by injecting 100 rats with a unit dose of the drug, subjecting each to neurological stimulus, and recording its response time. The neurologist knows that the mean response time for rats not injected with the drug is 1.2 seconds. The mean of the 100 injected rats' response times is 1.05 seconds with a sample standard deviation of 0.5 seconds. Do you think that the drug has an effect on response time?
H_0: μ=1.2 s; H_1: μ≠1.2 s.
σ_{x̄}≈0.5/√100=0.05; Z=(1.2-1.05)/0.05=3; P-value:.003.
The clip does not state the exact significance level chosen before the verbal 5% discussion; it only uses 5% as a common threshold later.
Decide whether a drug affects mean response time, given baseline mean 1.2 s, sample size 100, sample mean 1.05 s, and sample standard deviation 0.5 s.
n=100 injected rats.
Baseline mean without drug = 1.2 seconds.
Sample mean of injected rats = 1.05 seconds.
Sample standard deviation = 0.5 seconds.
Question asks whether the drug has an effect.
Test H_0: μ=1.2 against H_1: μ≠1.2 and decide whether to reject H_0.
Set up a two-sided test about the population mean response time under drug treatment.
Written directly on the board.
Compute the standard error of the sample mean.
Displayed formula and arithmetic on the right side of the board.
Standardize the observed sample mean relative to the null-hypothesis mean.
Displayed formula under 'Assume H_0'.
Read the combined tail probability from the normal-curve sketch and identify it as the p-value.
Diagram labels 0.3% and .003; audio defines this as the p-value.
Compare the p-value with the mentioned 5% threshold and make the decision.
Spoken rule and explicit conclusion in the audio.
Reject the null hypothesis; the data support that the drug has an effect on response time.
The conclusion follows from the displayed p-value 0.003 being far below the verbally stated 5% rejection threshold.
A black digital board shows typed white problem text across the top throughout the clip.
Below it, yellow handwriting adds H_0 and its symbolic form; green handwriting adds H_1 and its symbolic form; yellow handwriting later adds 'Assume H_0'.
Typed problem statement at top
Yellow H_0 line
Green H_1 line
Yellow 'Assume H_0' line
Cursor/pen strokes during writing
First the problem statement is read while the lower board is blank.
Then H_0 is written in yellow with the verbal phrase and symbolic equation.
Then H_1 is written in green below it with the verbal phrase and symbolic inequality.
Finally the next step 'Assume H_0' begins in yellow near the bottom.
The original typed problem text stays visible throughout.
The numerical values 100, 1.2, 1.05, and 0.5 do not change.
The parameter consistently denotes the mean response time with the drug.
The visual organization mirrors the logical sequence of the lesson: state the empirical problem, formalize the null and alternative hypotheses, then begin the inferential procedure by assuming the null.
A horizontal axis is drawn, then a symmetric bell-shaped curve above it, then a dashed vertical line through the peak.
The center is labeled μ_{} = μ = 1.2 s.
The speaker says this is the sampling distribution if we assume the null hypothesis.
Horizontal axis
Bell-shaped curve
Dashed vertical center line
Label μ_{} = μ = 1.2 s
First the axis is drawn.
Then the bell curve is sketched above it.
Then the center line is added.
Then the center label is written beneath the curve.
The curve is symmetric about the dashed center line.
The center corresponds to the null-hypothesis mean 1.2 s.
The visual represents the sampling distribution of the sample mean under H_0, centered at 1.2 seconds and treated as approximately normal because the sample size is 100.
To the right of the curve, the board writes σ_{} = σ / √100, then ≈ s / √100, then = 0.5 / √100 = 0.5 / 10 = 0.05.
The speaker narrates the substitution of s for σ and the arithmetic simplification.
σ_{} expression
s / √100 approximation
= 0.05 result
The symbolic formula appears first.
It is rewritten with s in place of σ.
Numerical substitution follows.
The square root is simplified to 10.
The final decimal 0.05 is written.
All expressions refer to the standard deviation of the sampling distribution of the sample mean.
The denominator remains √100 until simplified.
The visual progression shows how the unknown-population formula becomes an estimated numerical standard error for the example.
A yellow bell curve is drawn with center labeled μ_x̄ = μ = 1.2 s.
Yellow bell curve
Center label μ_x̄ = μ = 1.2 s
Assume H_0 text
The curve is introduced as the distribution assumed under H_0.
The center remains at 1.2 seconds throughout the clip.
The picture represents the sampling distribution of the sample mean if the drug has no effect.
Orange writing adds Z = (1.2 - 1.05)/0.05 and then Z = .15/.05 = 3.
Orange Z equation
Numerator 1.2 - 1.05
Denominator 0.05
Result 3
The expression is built step by step from the general ratio to the simplified numerical result.
The same three quantities 1.2, 1.05, and 0.05 remain in the calculation.
The visual sequence shows standardization of the observed sample mean relative to the hypothesized mean.
Vertical dashed lines mark 1, 2, and 3 standard deviations on both sides of the mean.
Speaker counts one, two, three standard deviations in the positive and negative directions.
Dashed vertical lines
Left and right sides of the bell curve
Mean line
Lines are added sequentially outward from the center to ±1, ±2, and ±3 standard deviations.
Spacing is presented as equal standard-deviation intervals around the mean.
The marks convert the abstract z-value into a spatial location on the sampling distribution.
The observed result 1.05 is placed at the third negative standard-deviation line.
Speaker says the 100 rat sample result is right over here, three standard deviations below the mean.
Observed value 1.05
Third negative standard-deviation line
The computed z-value is translated into a point on the curve.
The point stays on the left side because 1.05 < 1.2.
The sample mean is visualized as an extreme low value under H_0.
Both far tails beyond ±3 standard deviations are shaded magenta.
Speaker describes results less than this or that extreme in the positive direction.
Left tail beyond −3 SD
Right tail beyond +3 SD
Magenta shading
The focus expands from one observed tail to both symmetric tails.
The cutoff remains at three standard deviations from the mean.
Because H_1 is μ ≠ 1.2, extremeness is measured in either direction.
Central region is shaded reddish-orange and labeled 99.7%.
Speaker says 99.7% of the probability is within three standard deviations.
Central shaded region
Label 99.7%
The middle area is highlighted to separate it from the two tails.
The central interval is from −3 to +3 standard deviations.
The visual label supports computing the remaining tail probability by complement.
Bell-shaped green curve centered at μ_{x̄}=μ=1.2s with orange hatching over the middle labeled 99.7% and purple/pink tails labeled 0.3% and .003.
Around 10-28 seconds, arrows and labels are added above the tails to indicate the combined tail area.
The exact color naming varies between purple and pink in speech versus appearance; the mathematical role of the tails is clear.
Green bell curve.
Center label μ_{x̄}=μ=1.2s.
Orange central shaded region labeled 99.7%.
Two tail regions labeled 0.3% and .003.
Curved arrows connecting the tails.
Tail annotations are emphasized after the central 99.7% region is already visible.
The decimal form .003 is written next to the percentage 0.3%.
Later the board adds P-value : .003 at the bottom.
The center remains fixed at 1.2 seconds throughout the clip.
The total probability interpretation remains that central area plus tail areas partition the curve.
The example remains two-sided, with both tails contributing to the extremeness criterion.
The picture visualizes that an observed sample mean 3 standard errors from the null center falls into a region whose combined tail probability is only 0.003, motivating rejection of H_0.
At about 60-67 seconds, 'reject' is written near H_0.
At about 100-112 seconds, 'P-value :' is written and then '.003' is appended.
At about 141-146 seconds, H_1 is circled.
H_0 line.
H_1 line.
Bottom-left writing area.
P-value label.
The word 'reject' is added beside H_0.
The term 'P-value' is introduced in writing.
The numeric p-value .003 is written after the label.
H_1 is circled at the end to mark the favored hypothesis.
The earlier formulas for standard error and Z remain visible.
The central curve and tail labels remain on screen while conclusions are added.
The writing sequence mirrors the reasoning order: compute evidence, name the tail probability as a p-value, then record the decision to reject H_0 and favor H_1.
The speaker says that if the probability under the null is really small, then the null probably isn't true and we could reject it.
This caution is added by the analyst and is not explicitly stated as a misconception warning in the clip.
One might think that rejecting H_0 proves the alternative hypothesis is definitely true.
The clip only supports a probabilistic inference: if the observed result is very unlikely under H_0, that gives grounds to doubt H_0. It does not establish H_1 with certainty.
The board writes H_1: 1.2 s.
The clip does not explicitly discuss one-sided versus two-sided tests; this clarification is analyst-added.
One might interpret 'the drug has an effect' as automatically meaning it decreases response time because the sample mean is 1.05 s.
The written alternative is 1.2 s, so the hypothesis being tested is any departure from 1.2 seconds, not only a decrease.
The speaker says, 'We do not know what the standard deviation of the entire population is. So what we're going to do is estimate it with our sample standard deviation.'
One might think the formula σ_{} = σ / √n requires a known numerical σ from the whole population.
In this example σ is unknown, so the lesson explicitly replaces it with the sample standard deviation s and marks the result with a hat to show it is an estimate.
The speaker says, 'what we're going to do is not just figure out the probability of this, the probability of getting something like this or even more extreme than this.'
A learner might think the test asks only for the probability of getting exactly 1.05 seconds.
The video states that the relevant probability concerns results like the observed one or more extreme, which is the usual tail-area framing used before introducing a p-value.
Speaker writes 1.2 - 1.05 just so that it'll be a positive distance, then later says the result is three standard deviations below the mean.
The observed value is placed on the left side of the curve.
Because the board shows 1.2 − 1.05 and obtains +3, one might think the sample lies above the mean.
The video uses the subtraction order to display a positive distance, but the interpretation on the curve is explicit: 1.05 is below 1.2, so the sample is three standard deviations below the mean.
When I talk about this extreme, it could be either a result less than this or a result that extreme in the positive direction, more than three standard deviations.
Both tails are shaded.
One might compute only the left-tail probability because the observed sample is low.
Since H_1 is μ ≠ 1.2, the event 'this extreme' includes both tails beyond ±3 standard deviations.
I don't know 100% sure, but if the null hypothesis was true, there's only a 1 in 300 chance of getting this.
...this is a very strong indicator that the null hypothesis is incorrect and the drug definitely has some effect.
The speaker uses the word 'definitely' colloquially while also saying he is not 100% sure; the mathematical content still frames the result probabilistically.
One might read the conclusion as proving beyond all doubt that the drug works.
The video itself stresses uncertainty: the result is very unlikely if H_0 were true, which supports rejection, but it is not presented as 100% certainty.
H_1 uses μ≠1.2 s.
Both tails are marked and combined into 0.3% / .003.
One might count only the left tail because the observed mean 1.05 is below 1.2.
Because the alternative is μ≠1.2, the lesson combines both tails to get the total extreme probability 0.003.
The typed problem asks whether the drug has an effect.
The board converts that question into H_0: = 1.2 s.
The applied research question is translated into the formal null hypothesis that the drug-group mean equals the untreated mean.
H_0 and H_1 are written one above the other using the same parameter and the same reference value 1.2 s.
The speaker introduces them as two hypotheses for the same question.
The null asserts no effect via equality, while the alternative asserts an effect via inequality; they are competing statements about the same population mean.
After stating H_0 and H_1, the speaker explains how to decide between them by assuming H_0 and evaluating the sample result.
The board adds 'Assume H_0'.
The decision procedure depends on having already formulated both hypotheses; the next step is to test the observed data against the assumption that H_0 holds.
The speaker says to assume the null hypothesis and then think about the sampling distribution under that assumption.
The center label μ_{} = μ = 1.2 s is written after stating H_0.
The mean of the sampling distribution used in the test is taken directly from the null hypothesis value μ = 1.2 s.
The board moves from σ_{} = σ / √100 to ≈ s / √100.
The speaker explains replacing unknown σ with sample s.
The estimation method is applied to the general standard-error formula because the population standard deviation is unavailable.
The estimated value = 0.05 is written before the speaker asks how many standard deviations away 1.05 is.
The speaker identifies the next task as figuring out a z-score.
The clip does not complete the z-score arithmetic.
Computing the estimated standard error supplies the denominator needed to standardize the observed sample mean into a z-score.
The speaker asks for the probability of getting a result at least that many standard deviations away from the mean.
This connection is inferred from the stated goal and the video title context, not fully derived inside the clip.
Editorial note: in standard hypothesis testing, the z-score is the intermediate quantity used to compute the tail probability associated with the observed result under H_0.
H_0 gives μ = 1.2, which is then used in Z = (1.2 - 1.05)/0.05.
The hypothesized mean from H_0 is required to form the z-statistic.
σ̂_{x̄} = 0.05 appears as the denominator of Z.
The estimated standard deviation of the sample mean supplies the scale used in standardization.
After obtaining Z = 3, the speaker asks for the probability of a result this extreme and invokes the empirical rule.
The computed |Z| = 3 allows direct use of the empirical-rule statement for three standard deviations.
H_1 is μ ≠ 1.2 s.
Both tails beyond ±3 are shaded.
The two-sided alternative determines that extremeness is assessed in both directions.
Standard error 0.05 is used directly in the Z formula.
The Z-statistic cannot be computed in this example until the standard error of the sample mean is obtained.
Board shows H_0: Drug has no effect => = 1.2 s.
Board shows H_1: Drug has an effect => 1.2 s.
Speaker explains assuming H_0 and checking the probability of the observed sample result.
Handwriting begins 'Assume H_0'.
Speaker says, 'Let's assume that the null hypothesis is true.'
Typed text lists 100 rats, 1.2 seconds untreated mean, 1.05 seconds sample mean, and 0.5 seconds sample standard deviation.
The speaker says, 'if we assume the null hypothesis is true, let's try to figure out the probability...'
μ_{} = μ = 1.2 s is written under the bell curve.
σ_{} = σ / √100 is written on screen.
The speaker says σ is unknown and estimates it with the sample standard deviation.
The board shows 0.5 / √100 = 0.5 / 10 = 0.05.
The speaker says they are essentially figuring out a z-score.
The clip stops before the numeric z-score is calculated.
Speaker introduces z-score or z statistic and asks how far we are away from the mean.
Covered · Typed problem statement is read aloud and all numerical givens are introduced.
Covered · The speaker sets up the null hypothesis and writes it verbally and symbolically.
Covered · The speaker introduces the alternative hypothesis and writes the two-sided inequality.
Covered · The clip explains the logic of assuming H_0 and judging how surprising the sample result is; it ends just as 'Assume H_0' is written.
Covered · Opening board state and spoken setup of H_0, H_1, and the conditional probability question under the null.
Covered · Drawing and labeling the null-centered sampling distribution of the sample mean.
Covered · Writing the symbolic formula for the standard deviation of the sampling distribution.
Covered · Substituting s for σ and computing the estimated standard error 0.05.
Covered · Reframing the observed mean as a distance in standard-deviation units and identifying the next step as a z-score; no final z-value or p-value is reached in this clip.
Covered · Problem statement, hypotheses, and prewritten standard-error estimate are visible while the speaker transitions to computing the z-score.
Covered · The speaker writes Z = (1.2 - 1.05)/0.05 and explains using a positive distance and the 0.05 denominator.
Covered · The arithmetic is completed to Z = .15/.05 = 3 and interpreted as three standard deviations from the mean.
Covered · Integer standard-deviation marks are drawn on both sides of the bell curve.
Covered · The observed sample mean 1.05 is placed at the third negative standard-deviation position.
Covered · The speaker defines 'this extreme' for the two-sided alternative and shades both tails beyond ±3 standard deviations.
Covered · The empirical rule is invoked, the central region is labeled 99.7%, and the remaining two-tail probability is set up by complement.
Covered · Opening board state plus explanation that the two tails together are 0.3% = 0.003.
Covered · Speaker interprets the observed result as having only a 1-in-300 chance under H_0 and begins favoring the alternative.
Covered · Explicit naming of the p-value, writing P-value : .003, discussion of the 5% threshold, and final rejection of H_0.
Reviewed current material from 43 seconds states the null and two-sided alternative hypotheses, derives the standard error and z statistic, interprets the two tails as a p-value, and compares it with a 5% decision threshold; the sign-presentation limitation is disclosed.
YouTube2:21Prerequisite in Probability and statistics: Sampling → Hypothesis testing.
Sampling · View learning sequence
YouTube2:01Prerequisite in Probability and statistics: Distributions → Hypothesis testing.
Distributions · View learning sequence
YouTube2:38Prerequisite in Probability and statistics: Distributions → Hypothesis testing.
Distributions · View learning sequence