A common assignment on Quantitative Research in Communicationis as follows:
TASK A.
Association as defined in statistics is at least two variables that can be related. To associate two items, is to compare the items likeness or dis-likeness. Association is sometimes thought of as causation but has no implied weak likeness.
Identifying association is not a direct common link with two associated variables, therefore it is nonrandom, so it includes a small amount of an operative loss of defined order.
An association example is like comparing the amount of hot tea consumed by women to women becoming pregnant, which is more associated to the other in a positive manner. Other way to ask this question is which causes which; does drinking hot tea cause more women to become pregnant, or do more pregnant women drink hot tea? Both have nothing to do with each other, the increase in both are due to cold dark nights during the winter.
Test that are used to check association are the chi-squared test, P test, and, t-test.
A chi-squared test compares two associated objects for sampling similarities. Test inspects compared number of times with two different hypotheses that are not associated. Assuming any one of a set of values as independent then the expected number of times on a contingency table. Therefore if the value of chi-squared test being associated is too big, the deal that is seen could be different amount of compared numbers that will be low and a independent guess will be no good, then no association is rejected.
TASK B.
Causation is the act or process of causing. ( Merriam Webster’s ninth new collegiate dictionary, 1990) In statistics, change one value of a variable and this will cause the other value to change. When these values change, they could be a small correlation all the way down to zero. Two most common factors that influence causation are Cause and Effect. A concept of cause-and-effect is more contrived than the theory of common logic.
The following ideas are equivalent:
Plain English: If A then B., Formal logic: A ⇒ B … “A implies B” , Boolean algebra: B + ∼ A … “B or not A” Probability: P(B|A) … “probability of B given A”
The four ideas are not anywhere close to, “A causes B”. (Internet, 2007)
Example:
“A” equals person has malaria symptoms, “B” equals person is injected by a insect
If we were to make a spreadsheet that would have two rows for each person. And have two columns (A and B). In this spreadsheet we should see that A ⇒ B. This could be checked in all rows, that either have B or ∼ A; the, person was injected, or has no malaria symptoms. On the same spreadsheet, it would show that A did not cause B. The insect injection was not caused by a person’s malaria symptoms. Effect does not happen before cause.
If concluded in general, an ordinary collection of data should not be used to form a relationship for cause and effect. It can be suggested that A expresses indirectly to B, but not that A causes B. (Internet, 2007)
These two variables, A and B, could be seen as bad or unreliable data observation if only cause or effects are looked at.
TASK C.
Averages during computing correlation there could be problems with variability. Based to high on averages brought into action for lone groups. Covariance observed is as high as the possible covariance, the correlation of 1, showing very close simulate placement of the two equalized items. Co variation values that are negative could be as low as a perfect -1, equaling greatest plus equalized item of a value that has the most non-positive values with others. Variable correlated equalized items that indicates 0 are randomly related in the middle by two values. Correlation are values that can be interchanged without changing its values, this does help prove the flow of direction in causation. Other values depend on different variables; if there covariance share with other values in a correlation then this could be false.
So, statisticians should understand, no linear likenesses are being correlated for the two values. Statisticians need to be vigilant because correlation may have small amounts of errors, to include data replaced by a fake plain or ranges. Also the lack of having the same statistical deviation will in theory be true by a few percentages, that are with coefficients alone.
And here is what is wrong with the paper:
- The meaning of several intended points is unclear due to missing and misspelled words.
- The work contains an excessive number of mechanical errors. Please proofread it again. Please add full APA-formatted reference list citations for the in-text citations.
- The first sentence is a reasonable definition. The phrase “no implied weak likeness” is not understood.
- The definition relies heavily on the root word, cause. Please complete the definition by defining the root word.
Although it is difficult to follow, it appears the malaria example is suggesting a reverse cause and effect. This will be a good answer. Please discuss it more clearly. The specialty logic symbols do not come through, but they are not important to the explanation. The tea/pregnancy example in part A (suggesting a third variable) will work for the required second factor. Please incorporate that example in the answer to B.
Please explain what is meant by “problems with variability.” Also, the following sentence fragment is not understood: “Based to high on averages brought into action for lone groups.”
The intended points of the last paragraph are not understood. Please clarify. Be sure to discuss problems associated with the use of averages, in particular, rather than statistical analysis, in general. It would be appropriate to discuss the impact of an outlier or the problem of ecological fallacy.
