Clearly state the direction.

Can we explain the value of the sample mean by the standard deviation?

expected frequencies

Odds and odds ratio are always relative to second row/column.

telling us "How much the difference between those sample means is likely to vary from sample to sample."

The standard deviation of the difference is not the difference between the two given values of the standard deviation.

The standard deviation of the difference is not the difference between the two given values of the standard deviation.

Ecological validity: tells us whether the results are likely to be realised in the real world

Statistical validity: tells us whether our statistical methods look appropriate

tells if the conclusions we have made within our sample are reasonable

tells us if our sample represents the population

the sample standard deviation

the sample mean

the sample mean

the population mean

the standard deviation is σ

the mean of the distribution is µ

抽出変動、標本変動

A different sample would have produced different sample odds.

条件付き確率

Numerical summaries are difficult for describing the relationship between two quantitative variables, because the possible relationships vary greatly.

A correlation coefficient is a single number encapsulating all this information.

not necessarily the difference b/w these two individual Std.dev.s.

Std.dev. for all the individual reduction

numerical summary table

simplify with the straight lines indicating each range

a bit messy

a way to summarise quantitative data;

describes what values are present in the data, and how often those values appear

Understanding variables we have is quite important.

❓❓❓

because both types of variables can be summarized and analysed numerically

People may answer the questions even if they don't understand the questions.

leading to making a big difference to the ansers that we get

Complicated to answer

Asking two questions ... better to ask two questions here

Most people will give answers even if they actually don't know.

⇒ getting useless data

Ambiguous ... What do you mean? (more than yesterday? last week? WHEN?)

Leading ... very obvious from the way the question is posed what answer everyone is wxpecting

EXCLUSIVE

Make sure that they're exhaustive → cover all options

clarity in our questions → we get the answers we seek

Avoid people we're asking about a topic breaking laws, or revealing confidential or private information

two things in one question

Ensure people we're asking understand of what we're taking about.

respondents may be pushed in the direction

a set of questions for respondents to answer

What types of analysis is going to use?

What software we use for those analyses?

How we get the information from/about those indeviduals?

How we get the indeviduls from the population?

everybody can understand exactly what was done, and how

a small test conducted before the actual data collection, not whole test

a small test conducted before the actual data collection, not whole test

This leads to better future studies.

repeatable ⇒ enable others to replicate ⇒ can be ethical

a procedure documenting the details of the design and implementation of studies, and for data collection

a plan = should be established and documented before collecting the data → explains exactly how the data will be obtained, which will include operational definitions

a procedure documenting the details of the design and implementation of studies, and for data collection

effectiveness

generalizability

The likely practicality of the study results in the real world

検死

• by explaining some variation in the response variable.

by avoiding lurking variables (Sect. 3.4).

by determining if the comparison groups are similar (Sect. 7.2).

by using the information in analysis (Sect. 7.2).

If the researchers, participants and the analyst are blinded to the comparison groups, the study is called triple blind.

If both the researchers and participants are blinded to the comparison groups, the study is called double blind.

If only the individuals are blinded to the comparison groups, the study is called single blind.

individuals in a study may report effects of a treatment, even if they have not received an active treatment

暗算

∙ Suppose the researchers assessing the study outcomes knew the diet allocated to each patient.

∙ Researchers’ expectations or hopes for how the new diet will perform may unconsciously influence how the researchers interact with the individuals, and so perhaps (unconsciously) influence the behaviour of the individuals in the study.

白衣高血圧

temporary rise in blood pressure in the doctor's office.

∙ The impact of the Hawthorne effect can be minimized by blinding the individuals, so that:

the individuals do not know that they are participating in a study; the individuals do not know the aims of the study; and/or the individuals do not know which treatment they are receiving in the study.

∙ Blinding people to knowing they are involved in a study is often difficult, as ethics often requires peoples’ informed consent.

ホーソン効果

People, and perhaps animals, may behave differently if they know (or think) they are being watched

This is true for identified potential confounders (such as age), and also for variables not even considered as confounders, or are hard to measure or observe (such as genetic conditions).

→ manage lurking variables

recording all potential extraneous variables is important

The impact of some extraneous variables on the response variable can be reduced by fixing the values of the variable. = control variable

Control (or controlled) variables are extraneous variables whose values are fixed for the study.

There are many differences between two groups ... 👆 those differences written above

A confounding variable (or a confounder) is an extraneous variable associated with the response and explanatory variables.

Confounding is when a third variable influences the observed relationship between the response and explanatory variable.

they also changed the amount of labour and herbicide, not only the presence of fertilizer

Voluntary response (self-selecting) sample

non-random sampling

The sample may not be representative of the population for many reasons.

These compromise how well the sample represents the population (i.e., compromises external validity and accuracy).

trying to have a mix of students as a sample

Multistage sampling: larger collections of individuals are selected using a simple random sample.

Smaller collections of individuals within those large collections are selected using a simple random sample.

The simple random sampling continues for as many levels as necessary, until individuals are being selected (at random).

clusters

a large number of small groups

Cluster sampling: the population is split into a large number of small groups (clusters).

A simple random sample of clusters is selected, and every member of the chosen clusters become part of the sample.

The population is split into a small number of large (usually similar) groups called strata

Then cases are selected using a simple random sample from each stratum.

a small number of large (usually similar) groups

strata

the first case is randomly selected; then, more individuals are selected at regular intervals thereafter

a list of all members of the population

every possible sample of a given size has the same chance of being selected

maximising external validity

you have no say in what's going on in choosing the sample

Top Left: both Accuracy & Precision = very closely to the target & doesn't have so much variation around it

Bottom Left: Accuracy & Imprecision = hitting on the target on average, but not close to the target

Top Right: Inaccuracy & Precision = not much variation

Bottom Right: Inaccuracy & Imprecision = missing average, all the place

Many samples are possible, and every sample is likely to be different.

The results of studying a sample depend on which individuals are in the studied sample.

Many different samples are possible.

A study is externally valid if the results from the sample can be generalised to the population, which is only possible if the sample faithfully represents the population.

タスキギー梅毒事件、タスキギー人体実験

e.g. enginerering & chemical studies

decosion-making & relational

POCI = in Abstract

e.g. P: adults with type 2 diabetes

O response variables: glycated hemoglobin (AIC), subject's weight * two time periods (pre- and 6-month-post program)

C: 2 types of education being received (diabetes patient education OR augumanted by a community self-management program) = explanatory variables

I: 2 types of education being received *assigned

POCI ⇒ Inerventional RQ ⇒ experiment

Population, Outcome, Comparison and Intervention

Example 1: Unit of observation = each flower

Unit of analysis = each bunch = 6

Example 2: Unit of observation = each frozen cube = 24

Unit of analysis = carton = 2 (regular milk & chocolate milk)

Example 3: Unit of observation = individual students

Unit of analysis = individual students

units of analysis

one unit of observation

units of observation

unit of analysis

A two-tailed test, in statistics, is a method in which the critical area of a distribution is two-sided and tests whether a sample is greater than or less than a certain range of values. A two-tailed test will test both if the mean is significantly greater than x and if the mean significantly less than x.

A one-tailed test is a statistical test in which the critical area of a distribution is one-sided so that it is either greater than or less than a certain value, but not both.

explanatory = independent variables = explain changes in response

response = dependent variable = changed response to explanatory

Quantitative

Qualitative

Quantitative

Quantitative