Findings

What conclusions did the report come to, and are these supported by the data?

The findings are often hinted at in the start of a statistical report, and described in more detail at the end.

Sometimes there are findings that are not closely related to the purpose of the report. This could be because a lot of data was collected, and some unexpected patterns emerged.

Findings

The findings will be described with a statistical measure, or a comparison of statistical measures, for instance:

The probability of a certain outcome (or a comparison of the probability of the outcome between two groups); this could be how often a particular qualitative response was given
The mean or median of a quantitative variable (or a comparison of an average between two groups)

Claims

Sometimes a statistical report will make claims that cannot be supported by the data collected. Unfounded claims could include the following situations.

Causation

Only a controlled experiment can truly show that one thing causes another.

If a study finds that college students with good short-term memories report enjoying classical music, the report should NOT claim that classical music causes improved short-term memory in college students.

Transfer of Findings

The target population should not be broadened to make the scope of the study seem wider than it really was.

If a study finds that college students with good short-term memories report enjoying classical music, the report should NOT claim that this would be true of all people (beyond college students).

Misinterpretation

The wording of the findings should not be modified to change the focus of the study.

If a study finds that college students with good short-term memories report enjoying classical music, the report should NOT claim that enjoying classical music is related to college students being more likely to pass.

Graphical Representations

The way data is presented can make findings look more significant that they are. This could include:

poorly constructed axes, especially not showing zero
incorrect data displayed
3D charts, as areas are no longer proportional

Graphical representations also typical show means and proportions as if they are certain, with no indication of margin of error or uncertainty.

Title

The title of an article - especially a mainstream media article - could overstate the findings.

It might not be as dramatic as Buzzfeed clickbait, but media content is consumer driven, and they want our attention. The title might try to say more than the rest of the article.

(There's also a smaller report about Buzzfeed articles here: Reports).

Fish and chips are still top dog in New Zealand

31 Mar 2017, Carly Holloway, Nielsen (article link)

In an age where consumers say they are increasingly health aware, New Zealanders still regularly indulge in fast food. Research from Nielsen’s Consumer and Media Insights (CMI) survey reveals that in the past month, as many as 80% of New Zealanders ate fast food and more than a quarter (27%) ate fast food more than five times. Fish and Chips continues to be our fast favourite, with 1.7 million Kiwis eating it in the last month - an increase of 11% over two years.

Domino’s wins on growth. The number of people who have bought or eaten Domino’s in the last month has jumped by 39% compared to 2014. This is largely due to an increase in the number of teen customers, perhaps due to their new value positioning.

The data was gathered via an independently audited survey of 11,000 New Zealanders aged 10+.

164 words

Focus Questions

What are the overall findings of the report?
What further information is needed?
Are the claims made in the statistically based reports valid and/or sensible?

Possible Answers

The key findings of the report are that New Zealanders buy fish and chips more often than any type of takeaway. Domino's have improved their market share the most in the last two years.

The survey has asked which fast food was eaten in the last month, and used this to express a 'favourite'. It might be that some people have a favourite fast food, but only eat it infrequently due to price or accessibility. The claim that fish and chips is a 'fast favourite' cannot be justified.

The way the responses have been structured includes some brands (like McDonald's and Domino's) and some types (like Indian and Sushi). If the brands were put together as types, then Burgers would be the 'favourite' choice for New Zealand consumers, with fish and chips in second, and pizza in third.

Task 7

Answer the Findings focus questions about the Drink Driving report:

What are the overall findings of the report?
What further information is needed?
Are the claims made in the statistically based reports valid and/or sensible?

Report-Drink Driving.pdf

Google Sites

Report abuse