One way to assess the precision of a statistic (a point estimate) is to compute the standard error, which is the standard deviation of the statistic's sampling distribution. A relatively large standard error indicates that the point estimate should be viewed with skepticism, either because the sample size is small
Reading quite a lot of the press and commentary about analytics, you could be forgiven for thinking that these techniques were only usable for customer-facing work. Improving sales and marketing? Yes, you want analytics, using Big Data about your customers. But there are plenty of internal and operational issues where
Today in higher education, savvy users expect to have the information they need to make data-informed decisions at their fingertips. As such, leaders in institutional research (IR) are under pressure to provide these users with accurate data, reports and analyses. IR has been tasked with transforming data and reports in
There have been several polarizing topics throughout history, such as religion & political affiliation. And for software developers there's one more biggie ... tabs -vs- spaces! Which group is right? Perhaps the opinion of the better programmers should have more weight(?) Is there a metric we can use to determine whether
A Stem and Leaf plot is a visual that can help quickly visualize the distribution of the data. This graph was particularly useful before the advent of modern statistical graphs including the Histogram and Box Plot. One nice feature of the plot is it shows the actual values in the
Following the UK General Election result where no political party secured a clear majority ahead of the Brexit negotiations, it’s fair to say there’s lots of uncertainty facing the UK government right now. This, and the more outward looking post-Brexit era we're facing, are just two reasons why I believe
Most numerical optimization routines require that the user provides an initial guess for the solution. I have previously described a method for choosing an initial guess for an optimization, which works well for low-dimensional optimization problems. Recently a SAS programmer asked how to find an initial guess when there are
It has been almost a year since then-U.S. CIO Tony Scott introduced the federal open source policy that called for agencies to share federally-developed software source code. The policy, more than anything, aimed to make agencies more agile. Instead of redeveloping the same programs the open source policy would allow
A frequently requested statistical graph is the scatter plot by with discrete categories along with mean value for each category. Searching for a "Scatter with Mean" will return a lot of requests for such a graph in SAS, Stata, R and other statistical software. Such a graph is very easy
Have you ever written a macro and wondered if there was an easy way to pass values to the macro? You can by using macro parameters. Macro parameters enable you to pass values into the macro at macro invocation, and set default values for macro variables within the macro definition.
If countries have a similar median age, does that mean they are also similar in other ways? My best guess at an answer is - probably. Perhaps if we plot the data on a map, we'll be able to see the answer more clearly. I first started thinking about this
Regulations, corporate drivers, leadership and market influences have combined to produce a patchwork of uneven progress on initiatives such as distributed generation, customer choice, asset optimization and the industrial Internet of Things. These initiatives all rely on analytics to gain the most return on investment. To better understand organizational readiness
In part 2, Jim Harris explains more about why you should address data quality and governance issues on the way to data lakes and Hadoop.
When I was in college, I lived in the Alexander dorm, where each room had an American student and an International student. I learned a lot about the rest of the world, and I also heard a lot of different languages being spoken. I'm really impressed when I hear people
There's been a lot of talk in the media lately about the death of retail. Every week, it seems like another retailer announces the closing of stores, acquisitions or even going out of business. Many relate it to the growing competitive landscape with the convenience of online shopping and lure
With busy week nights, some foods don’t make it to the table until the weekend simply due to lack of prep time. But, it’s a shame to narrow some of our favorite foods to only a few nights of the week. So, a quick solution when you’re in time crunch
In a previous article, I showed two ways to define a log-likelihood function in SAS. This article shows two ways to compute maximum likelihood estimates (MLEs) in SAS: the nonlinear optimization subroutines in SAS/IML and the NLMIXED procedure in SAS/STAT. To illustrate these methods, I will use the same data
In SAS Viya 3.2, the Self-Service Import provides a mechanism for a user to import (copy) data into the SAS Cloud Analytic Services (CAS) environment. The data is copied as a .sashdat file into the selected CAS Library location when it is imported. Self-Service Import data can only be imported into
Customizing the Kaplan-Meier plot in assorted ways is so popular that we devote an entire chapter to it in the SAS/STAT documentation.
Maximum likelihood estimation (MLE) is a powerful statistical technique that uses optimization techniques to fit parametric models. The technique finds the parameters that are "most likely" to have produced the observed data. SAS provides many tools for nonlinear optimization, so often the hardest part of maximum likelihood is writing down
Registration is now open for the SAS Analytics Experience 2017, being held September 18-20, in Washington, DC. (The Analytics Experience moves to Amsterdam, October 16-18 -- details on that event to follow.) For anyone interested in FVA analysis, Chip Wells and I will be delivering a half-day pre-conference training session
If you were a fan of the original Star Trek television series, you probably remember lots of little details about the show. And you might even feel sorry for the people who don't get the clever references you make to things from the show. If you're that person, then you'll
On a recent trip I met a long time user and early adopter of ODS Graphics who started using GTL with SAS 9.1.3, even before it was released as production with SAS 9.2. This user has presented many papers at SGF on GTL and some hands-on sessions on ODS Graphics Designer.
Utilities can save as much as $10 million a year for every one percent improvement in forecast accuracy by optimizing asset utilization and trading strategies. Changing energy markets and an influx of data from the smart grid are providing more opportunities to reap value from energy forecasting. Improving energy forecasts
In the third post of the 10 Commandments of Applied Econometrics series we discussed the issue of data exploration. Today, I will present the next commandments: keep the solutions simple and use model validation. Keep the models sensibly simple Striking the right balance between simplicity and sophistication: the models created should be neither
"How'd you lose the weight?" It was not a magic pill, special diet, or a shot from the doctor - wouldn't that have been nice? How did I lose my baby weight? I finally got back to what I've been telling folks to do most of my life: eat right
Todd Wright says addressing GDPR compliance might be just the catalyst the insurance industry needs.
In a world that never stands still, especially in the post-Brexit whirlwind the UK will soon be entering, it’s imperative that government is agile and responsive. More importantly, how can this capability be arrived at without breaking the bank? The UK Government recently released its Transformation Strategy 2017 to 2020.
I have previously discussed how to define functions that safely evaluate their arguments and return a missing value if the argument is not in the domain of the function. The canonical example is the LOG function, which is defined only for positive arguments. For example, to evaluate the LOG function
