Testing a claim

What is a claim?

A statement that something is the case, typically without providing evidence or proof.

What is a sample?

a small part or quantity intended to show what the whole is like.

What is accuracy? Give an example of how something can be tested

Accuracy is the measure of how close a measurement is to the true value of that measurement.

\
A manufacturer claims that the desk has a length of 100cm.

To test this, an individual can measure the desk themselves using a measuring tape.

If the results obtained match with the claimed length then it is accurate.

\
(Another example is the cookie test done in class)

What is precision?

A measure of how close a series of measurements are to each other.

What is reliability?

An extent to which repeated observations and/or measurements taken under identical circumstances will yield similar results.

What is validity?

An extent to which tests measure what was intended, and to which data, inferences and actions produced from tests and other processes are accurate.

What is Bias?

inclination or prejudice for or against one person, group or a fact especially in a way considered to be unfair.

What is error?

Errors are effects that cause the measurements to differ from their true value.

What is random error?

Unpredictable fluctuations to measurements.

What is systematic error?

A mistake made when measuring for example, not calibrating or using the wrong equipment for the experiment.

What is contradictions?

Statement or depiction that is inconsistent and does not adheres to logic when compared to one or more other statements or depictions

What is sample bias?

The tendency to under or over represent a particular group within the sample population.

What is confidence level?

Confidence level is expressed as a percentage likelihood that a repeated investigation on a random sample will produce a comparable (similar) result.

What is confidence interval?

A value with an associated margin of error that indicates the range in which a random sample from the population is likely to fall into for a given measure or treatment.

What is stereotyping

The process of classifying someone (or something) on tje basis of superficial characteristics and attributing generalisations to all individuals that fit within the bounds of that classification

How can random errors be reduced?

Taking measurements to increase precision.

What can cause random errors?

It can caused by:

* Sudden changes in the experimental conditions that cannot be really controlled.

What can cause systematic errors?

It can be caused by:

* Malfunctioning in the equipment
* Incorrect calibration
* Incorrect observations- parallax errors

How can systematic errors be reduced?

* Calibrating the instrument before use
* Reading and using it correctly
* Peers checking for you
* Controlling things than may affect the measurement

What shapes a person in a way that leads them into obtaining bias results.

* Source of education
* Level of education
* Morals
* Beliefs
* Culture
* Demographics

What does reducing bias do?

Increases reliability and validity of the conclusions drawn

What are the 2 common ways in which sampling error can be introduced?

1. Inappropriate sample selection
2. Poor sample size

Is a big sample size important?

Big sample sizes eliminate random chance from swaying the results

What does small sample sizes cause?

The sample size being too small will cause random chance from swaying the results

Give an example of the effects of sample size

* If a coin is tossed 10 times to determine the chances of it landing heads or tails, random chance may produce a result that sees the coin land eight times on heads and only two on tails.
* From this, we could easily conclude that when a coin is tossed there is an 80 percent chance that it will land on heads.
* However, if we increase the sample size to 1000 tosses instead of 10, it becomes far more likely that we will see a more reliable results of a 50 percent chance of getting a head or tail.

What are the limits of collecting a large sample size?

* Costy
* Time consuming

How is the optimum sample size determined?

The optimum sample size is determined based on the confidence level and the confidence interval.

What is the key distinguishing factor of confidence interval and level?

\
Confidence level refers to the percentage of probability, or certainty, that the confidence interval would contain the true population parameter when you draw a random sample many times.

what is usually used to determine how large a sample needs to be to meet required reliability thresholds.

A sample size calculator

Give an example of a sample bias.

* Ashish wants to investigate the inquiry question:

‘How does the concentration of salt in the soil affect the growth of wheat plant?’

* Ashish gathers 20 large planting trays, fills them with identical soil and sets about planting out the trays with 20 wheat seeds in each before watering with various concentrations of salt water.
* Ashish has inadvertently selected a variety of wheat that has been bred to be salt resistant.
* This is an example of convenience sampling

Give an example of voluntary response samples.

Middle class people who speak english as their first language.

How can readers of scientific research understand the information in an incorrect manner?

* Due to translation errors
* The data being misinterpreted
* Innate bias and ideology of the reader clashing with the information presented
* Terminologies losing its intended definition due to them being translated into common (everyday) language. An example is the word ‘control’ even tho in a scientific perspective it insinuates a control group, others might perceive it as a group who control the experiment.

What are the 2 bias sampling?

1. Convenience sampling
* People who are easy to reach are picked for sampling
* People are closer have a relationship with the investigator are used as samples
* Not everyone has a chance to participate in the Sample
2. Volunteer sampling
* People volunteer
* Consist of people that have chosen to include themselves
* The people are biased towards the topic that they volunteer
* Not the entire population is included

What are the 5 non-biased samplings?

1. Simple Random sampling (SRS)
* Each individual has a chance of being surveyed
* Individuals are randomly selected
* Random digits table used
2. Systematic sampling
* Participants are selected at regular intervals (every 4th person)
3. Target population
* The entire population is used as samples
4. Stratified random sampling
* The population is divided in random samples
* Strata are groups of people with similar situations
* Equal amount of people are randomly selected
* Random digits table used
5. Multistage sampling

To or more SRS conducted


1. Simple random sampling is used to pick a group from the population
2. Simple random sampling is used to pick random people from the groups
3. Random digits table used