Interpreting Tables Flashcards
What does statistics mean?
The science of learning from data.
Which are the 2 types of data sources?
- Primary data.
2. Secondary data.
What are the primary data?
Collected specifically of investigation. Experiments. Observation. Interviews. Surveys.
What are the secondary data?
Already compiled. Available for analysis. Records. Databases. Health data.
What does the frequency in a table mean?
The cunt in each brand category.
How can we analyse qualitative data?
Ask participants a common questions.
Make a table base on the question, the participants, and their answers.
Calculate the frequency of their answers based on the question.
Calculate the %frequency.
Cumulative frequency.
% Cumulative frequency.
What is the cumulative frequency when calculated in a table of dataset took from participants for a qualitative data analysis?
Running total of the frequency column.
What is the % cumulative frequency?
Running total of the % frequency column.
What do we find when they ask: ‘What % of respondents did an action?’?
% frequency value.
What do we answer when they ask :’How many respondents bought this no more than twice?’?
Count frequency values of nil (o) + once + twice.
What do we answer when they ask: ‘How many respondents bought this at least 3 times?’?
Count frequency values of 4 + 4 + 5 + 6 + 7 + more.
What is the cumulative frequency?
The number of individuals/items in all categories up to and including the considered category.
What is the cumulative percent?
The cumulative frequency of a categories divided by the sample size multiply by 100%.
How can we design a questionnaire when the question is:
‘Who buys fast food most often?’
and we need to find:
the association between age group and no. of purchases?
Ask respondents their age and how often they buy fast food.
Pu data on a table.
Count total of data.
Where do we use contingency tables?
For nominal.
Ordinal variables.
Why do we use contingency tables?
To record.
Analyse relationship between 2/more variables.
What does each cell on the dataset represent?
The number of items into corresponding categories.
Where do we have totals on a dataset table?
Horizontally and landscape.
Which is a higher outcome?:
7 out of 8 people
or
7 out of 12 people?
7 out of 8 people.
What do we use when the counts are not very helpful to make them more informative?
Percentage.
Where do use percentage in contingency tables?
For nominal and ordinary variables.
How are the column percentages obtained?
(Each cell / corresponding column total ) * 100%.
How can we calculate row percentages?
(Each cell / corresponding row total ) * 100%.
How can we calculate total percentages?
(Each cell / sample) * 100%.
Whatis percentage equal to?
Proportion.
Probability.
If we have a SPSS Contingency table of household income v Internet use of participants.
How much is the % of internet non-users?
How much is the proportion of them?
How much is the probability of them in the sample?
Total amount of internet non-users / Total amount of all participants
* 100=
40%.
40%.
40% or P(internet non user) = 0.4
How do we show the probability value of a fact in the sample?
P(X) = value.
How much is the % of internet non-users with an upper household income?
Value of no-user upper / total non-users value * 100% = value.
Same with proportion and probability.
How do we show the probability of a fact with another fact in the sample, known as condition probability?
P(X|Y) = probability of X given the condition Y.
What will the probability of a randomly selected participant being a male who smokes be?
Male value / total number of participants.
What are the marginal tools?
Marginal probabilities.
Total numbers of questions asked to participants.
What is the probability of a randomly selected individual being a male?
Total male / total.
Not mentioned smoking/not.
Includes all cases.
What is the probability of a randomly selected individual smoking?
total who smoke/total.
No gender mentioned.
What is the probability of a randomly selected male smoking?
Males smoke / total males.
Only interested in male.
Only smokers.
What is the probability that a randomly selected smoker is male?
Male smokers / total smokers.
Focus on smokers.
Male only.
What are the comparing proportions?
Probabilities which provide evidence for/against associations.
What is the probability that a woman has breast cancer?
P(D) = Total with disease/Total in study.
What is the probability that a woman has breast cancer given she has taken HRT?
P(D | E) = no. taking HRT with disease / no. taking HRT.
What is the probability that a woman has breast cancer given she has not taken HRT?
P(D | not E) = no. of taking HRT with disease / no. not taking HRT in study.
What is the association between the values that a woman has breast cancer with and without HRT?
HRT taken –> bigger probability to have breast cancer = 0.01 > 0.007.
What % of patients had an outcome of ‘Full Recovery’?
Total number of full recovery / Total number of all.
Of patients who received a high dose, what % recovered fully?
High dose fully recover value / Total of high dose value.
Of patients who made a full recovery, what % received a medium dose?
Medium dose dully recover / Total of fully recover.
Of patients who died, what % received at least a medium dose?
Sum of medium dose value + high dose / Total of death value.
For patients who received a low dose of drug, what % of patients had at most a minor disability?
Sum values of low dose of death + vegetative state + major disability + minor disability / total of low dose value.
For patients who received at least a low dose of drug, what % of patients had at most a minor disability?
Sum of values 1, 2, .., until minor disability of low dose and more people /sum of low dose and more of patients.