Posts Tagged ‘statistics in the news’

When “Big Data” gets it wrong

January 4, 2017

Big Data could provide a big advantage to investors, but when the data is wrong or misinterpreted, it can be catastrophic.

“Credit card data sold to investors is making shares of retailers behave strangely, especially when the data gets things wrong.”  So begins an article in the Wall Street Journal ‘Big Data Adds Up to Trading Distortions.’  The first example is about Tailored Brands (owner of Men’s Warehouse and Joseph A. Bank) stock that shot up nearly 40% in one day when investors realized that the data they were basing their decisions on was inaccurate.

For the full article:  how-credit-card-data-might-be-distorting-retail-stocks-wsj

Scientists’ comprehension of confidence intervals is poor.

July 6, 2014

So much of scientific research involves statistical analysis and conclusions that have to be understood in the context of confidence intervals and margins of error. If our scientists don’t comprehend these concepts well, findings and reports are likely to be very misleading or wrong.

A study shows that this is the case.

https://mrsiderer.files.wordpress.com/2014/07/confidence-interval-comprehension.doc

At Proctor and Gamble Toothpaste is Data

September 8, 2011

This article is from Forbes magazine website.

TECH | 8/03/2011 @ 11:46AM |3,902 views

At Procter & Gamble, Toothpaste Is Data

Every Monday Procter & Gamble’s leadership gathers around an oval table on the sixth floor of the company’s Cincinnati, Ohio headquarters. The room is lined by two enormous and curved screens. This is the soapmaker’s “Business Sphere” and, like a philosopher’s stone, it provides a view into the 4 billion times each day that P&G products are used in more than 80 countries. In four clicks on an iPad it’s possible to change the screen from a view of a world map to a graph of toothpaste prices in India and a sales comparison that shows local brands are gaining share. Three clicks on a laptop bring up a view on another screen of shampoo sales in Australia. It’s immediately apparent that P&G will need to sell 585,000 more cases of shampoo for sales to be flat. On one screen indicators suggest big opportunities for hair care in Germany and pet food in the U.S.

The Business Sphere was turned on earlier this year. It’s the sum of 14 different technologies from multiple vendors, and from the first day P&G’s leaders began using it the data shown in this room have become the same data used by managers in some 40 locations worldwide (just not as fancily displayed). Aligning these numbers means that everyone knows and can agree on target countries, regions or products for particular attention. Everyone can judge progress toward the plan and against the competition, in some cases even down to individual retailers.

Procter & Gamble, the world’s biggest consumer products company, continues to be one of the leaders in the race to harness massive streams of data for managing a business better. The company has been profit-forecasting on a monthly basis for about 40 years, trying to predict components such as sales, commodity prices and exchange rates. But the amount of real-time data it has been able to process has increased vastly in the past three years, thanks to better software and Moore’s Law of increasing computing power. Now P&G borrows liberally from tools born on the Web: ubiquitous high-speed networking, data visualization and high-speed analysis on multiple streams of information. The tools allow P&G to make in minutes the decisions that used to take weeks or months, when data had to be collated and passed through committees on their way to the top.

With $79 billion in sales and 127,000 employees, P&G is on the verge of having everyone’s talents known and tracked, all information about sales decided at the executive level every week and production viewed in near real-time worldwide. The company talks in terms of increasing the amount of collected data sevenfold. The airy promises of networked technology are here, at a scale rarely, if ever, deployed before.

P&G has just started a “digital skills” inventory of its employees, establishing a baseline of skills, including how to get connected to the Internet, how to use basic collaboration and knowledge-exchanging tools for online meetings and mail, and how to tap into the company’s internal social network, P&G Pulse, for news and further training. There are higher expectations for more technical jobs, of course, or in grooming certain careers.

In one early pilot the P&G tech team figured that field salespeople gained 1.5% in revenue if they came to the customer with an iPad that could show the layout of different floor displays. Similar machines on the factory floor, able to take pictures of problems and replace paper filing systems with searchable databases, improved productivity by up to 20%. There are mobile apps for top managers to view reports, approve spending and get an alert when something goes off the rails, wherever they are.

The push is part of Chief Executive Robert McDonald’s goal to make the company fully digital. In an internal video introducing the initiative, McDonald walks his colleagues through his office, opening desk drawers to show he works without paper. “The idea is to be digital from the laboratory molecule to the store shelf,” he says.

Of course the quality of data can vary. “I used to work in China, and the one thing you knew when someone said, ‘We have a 22.5% market share’ was that the figure wasn’t right,” says Jeffrey Goldman, the statistician who sits in at the weekly Business Sphere gathering. “But it gets better over time, and it isn’t massaged on the way up.”