Question

What can improve representativity of a sample( more answers can be choosen)
1. adaptive sampling based on a priory knowledge about population
2. bigger sample size
3. smaller sample size

which variables are discrete but not ordinal qualitative (several can be choosen)
1. number of children in a family
2. height of people defined as ' small', 'average', 'high'
3. colors of hairs defined as 'dark' and 'light'
4. number of students attended a seminar

in which of the cases is the use of inferential statistics necessary in order to get answer
1 predict who will win presidential elections in USA
2. find average weight, height and IQ of students in class
3. investigate if a medicine helps to cure a wide spread disease
4. investigate if a medicine helps to cure a very rare disease
5. calculate average body temp of patients in a hospital

Answer 1

What can improve representativity of a sample

1. adaptive sampling based on a priory knowledge about population

because using this technique we can choose the sample based on another variable which possess some prior knowledge of the population. This works as an advantage over purely random sample as having prior knowledge to the population helps to understand the population structure. Hence can actually draw a better sample.

2. Bigger sample size

The more the sample at hand better is the estimation as in limiting sense, increasing sample size actually leads to acquire the whole population. Hence improves representativity.

which variables are discrete but not ordinal qualitative

1. number of children in a family.

It can take non negative integer values. hence discrete and clearly not qualitative as it can take numeric values

3. colors of hair defined as dark and light.

It is discrete as it can either take the value dark or light. and clearly not ordinal as we can not actually define an order to dark and light

4. number of students attended a seminar.

It can take non negative integer values. hence discrete and clearly not qualitative as it can take numeric values

in which of the cases is the use of inferential statistics necessary in order to get answer

1. predict who will win presidential elections in USA

It clearly involves inferential statistics as the method that needs to be used to answer this question involves maximum likelihood estimation.

3. investigate if a medicine helps to cure a wide spread disease

It also clearly involves inferential statistics as here we need to do testing of hythopethsis with null hypothesis be the persons affected with the disease and took the medicine got cured.