Handling Data Coursework
Presentation on theme: "HANDLING DATA COURSEWORK"— Presentation transcript:
1 HANDLING DATA COURSEWORK
2 Main Menu THE IMPORTANT STUFF THINGS YOU NEED TO KNOW
What is Coursework???Specify and PlanCollect, Process & RepresentInterpret and DiscussWhat You Should Do?THINGS YOU NEED TO KNOWPlanning the InvestigationSampleMean, Median,Mode and RangePie ChartsBar ChartsHistograms and Freq PolygonsScatter PlotsStem and Leaf PlotsCumulative FrequencyBox and Whisker Plots
3 “investigate what influences the amount a student drinks.”
Your TaskIs given in detail on the task sheet.Basically your task is to:“investigate what influences the amount a student drinks.”The database has been selected for you from Rondam Secondary school.
4 What Will HappenA MIX OF THE FOLLOWING:Direct Teaching – statistics skills, ICT, investigation cycleGroup Work – planning, discussing, plagiarism?Individual Time – writing up, working
5 Specify and Plan Specify and plan Hypothesis Interpret and discuss
How could you make it better?Interpret and discussSpecify and planInvestigation cycleCollect, process and represent
6 What to do in this section?
Examine the Writing Frame and what decisions you must make to fill it in.Decide on the hypothesis you are going to test. Make sure it is well explained.Write a clear and detailed description of the task and your plan to test the hypothesis.Do a draft first. Your final write up will come later.
7 Collect, Process and Represent
Specify and PlanCollect, Process and RepresentHypothesisHow could you make it better?Interpret and discussInvestigation cycleSpecify and planCollect, process and represent
8 What to do in this section?
Collect the data – fully explain your sampling technique and sample size.Tabulate the data. Only include the information relevant to your hypothesis.Using statistical and graphical methods to process and examine the data.
9 Specify and Plan Interpret and Discuss Interpret and discuss
HypothesisHow could you make it better?Interpret and discussInvestigation cycleSpecify and planCollect, process and represent
10 What to do in this section?
This is the big crunch section.Draw conclusions from all of your calculations and relate these to your initial hypothesis.Make sure you:Compare results to show differences/similarities.Use facts and statistics taken directly from your calculations.Evaluate your approach and explain any changes you would make if you were doing it again.Consider bias in your results.
11 And Now ……. Challenge 15 mins in groups of 5 or 6
What will a good piece of maths investigative work look like ???You should consider:What will it contain?How will it be presented?How will it be marked?What will it look like?15 mins in groups of 5 or 6
12 Formulating a hypothesis
The first step in planning a statistical enquiry is to decide what problem you want to explore.This can be done by asking questions that you want your data to answer and by stating a hypothesis.A hypothesis is a statement that you believe to be true but that you have not yet tested.The plural of hypothesis is hypotheses.Discuss the first steps in planning a statistical enquiry.Define a hypothesis as a statement of something that you believe to be true but do not have any evidence to support. For example, we could hypothesize that tabloid newspapers use shorter words than broadsheet newspapers.Year Eleven pupils with paid jobs don’t do as well in their exams.For example,
13 “Year Eleven pupils with paid jobs don’t do as well in their exams.”
Forming a hypothesis“Year Eleven pupils with paid jobsdon’t do as well in their exams.”How could you find out if this statement is true?Think about:What data (information) would you need to collect?How will you collect it?Which Year Elevens does this statement cover?How could you ensure the data you collect represents allof these Year Elevens?These questions could be discussed in groups, with each group feeding back their ideas to the class.While pupils discuss the question, issues of population, sample size and bias should arise, and you should introduce the vocabulary as it is needed. Definitions are summarized on the next two slides.Groups may decide that the population referred to in the statement is Year Elevens in their school, their area or region or the whole country.The sample should include pupils of all ability levels and equal numbers of male and female pupils (unless there is a good reason not to).Ethical issues should also arise. Some interviewees may not wish to disclose the number of hours they work; other pupils may have to work at home unpaid acting as carers. Some groups many decide to widen the statement to include unpaid work. However, people are not always good at estimating lengths of time and may exaggerate the number of hours they spend on housework, particularly if it is done unwillingly!Sample size should be taken seriously: if pupils are collecting their own data, a sample size of between 30 and 50 is probably realistic. (When using secondary data, such as information from the internet, books or school records, a larger sample is preferable.)When talking about how to make sure a sample is fair, it can be helpful to compare taking a sample to eating a slice of a pie: you don’t need to eat the whole pie to find out what it tastes like. So long as your slice includes the rim, the crust and the filling, you should get a good idea from that.Pupils are often vague about what they will do with the data once collected, and so often collect inappropriate data. For example, groups which decide to ask pupils “Do you have a job?” will have less useful data than those who ask “How many hours of paid work do you do?” They will also need to think about how to measure exam performance. Average point scores may be useful here (1 = G, 2 = F, etc).What would you do with the data?What would you expect to find?
14 Key vocabulary hypothesis – a statement that can be tested
population – the group (often of people) referred to in the hypothesissample – a selection from the populationbiased sample – an unfair selectionrepresentative sample – a fair selectioncross section – a selection that reflects all the subgroups within the populationAll of these words should be introduced in context, for example within the discussion on the previous slide. It is useful to give pupils a sentence using each word to model its correct usage.objective data – information that is not affected by people’s opinions
15 Key vocabularysubjective data – information that is affected by people’s opinionsprimary data – information you collect yourself, by asking people, measuring, carrying out experiments, and so onsecondary data – information that has been collected already, that you get from books, the internet, and so onethical issues – problems to do with confidentiality and personal questionsreliable results – results that will be repeated if the experiment or survey is carried out again with a new sample
16 Extending a hypothesis
Once you have collected data and drawn conclusions about your hypothesis, you could ask further questions and pursue other lines of enquiry.You will need to plan what these might be beforehand if you are carrying out a survey. For example,“People feel stressed when they have exams.”“You get less work done when it is noisy.”“Sleep deprivation affects concentration.”“Coffee can help you revise better.”These hypotheses have already been discussed on a previous slide, which will help to focus pupils on thinking only about extension questions rather than the many other issues involved in investigating them. This discussion is suitable for paired or group work.Suggestions for extensions might include:When else do people feel stressed? Is it before the exam more than during? Are stress levels reduced for certain types of people, such as organised people, confident people or older people?Do different types of noise have different effects, such as music with and without words, talking, outside activities such as drilling or lawn mowing?Does one night of sleep deprivation have less effect than several? Does it depend on how much sleep you need? Do people differ in how how much they need? Does this change with age? Does it depend on what you are concentrating on?What about other caffeine drinks, like tea or Coke? What about water, which is supposed to aid learning?Does it depend on when you revise (e.g. the night before or over a period of a week)? How could you measure the quality of the revision, rather than just its quantity? Does it depend on what resources you are using (e.g. websites, books, revision guides)? Does it depend on the subject or the teacher? Do people in the top sets revise more or less than those in the lower sets? Do boys revise more than girls?“The more revision you do, the better your exam results.”How could you extend these hypotheses?What extra information might it be worth collecting?
17 How are TV viewing figures compiled?
Sampling – Soap WarsDiscuss how pupils think television viewing figures are compiled.How are TV viewing figures compiled?
18 Television viewing figures
When compiling television viewing figures, it is impractical to find out what everyone in the country is watching at a particular time.Instead, the viewing habits of a sample of households is carefully monitored and the data collected is used to compile the figures.To avoid bias, it is important that the sample is representative of all television viewing households across the country.Discuss what categories of households would be important when monitoring viewing habits. For example, geographical location and household income levels.This is done by dividing households into categories and taking samples in proportion to the size of each category.This is an example of a stratified sample.
19 Different sampling methods
Random samplingPeople are chosen at random e.g. names picked from a hat or using a random number generator on a calculator.Every member of the population has an equalchance of being chosen.27Systematic samplingMembers of the population are chosen at regular intervals, such as every 100th person from a telephone directory.These kind of methods are used in research in contexts such as psychology, sociology, social policy, medicine, marketing, politics, economics. It will be useful to relate them to pupils’ own interests and future career plans.Quota samplingYou keep asking until you have enough people from each category. An example would be a survey in the street where you stop when you have enough people from each age category.
20 Evaluating different sampling methods
Random sampling Every member of the population has an equal chance ofbeing chosen, which makes it fair. It can be very time consuming and usually impractical.Systematic sampling You are unlikely to get a biased sample. It is not strictly random: some members of the population cannot be chosen once you have decided where to start on the list.These advantages and disadvantages are most useful in discussions in the context of particular hypotheses. Evaluating the choice of sampling method is a higher order thinking skill which is most appropriate to pupils aiming for A and A* grades. Ask pupils to come up with their own advantages and disadvantages before displaying these.
21 Evaluating different sampling methods
Quota sampling This is easier to manage. It could be biased. For example, if you are only askingpeople on the street or in a shop, the sample might notrepresent people at work all day.Stratified sampling It is the best way to reflect the population accurately.Ask pupils to come up with their own advantages and disadvantages before displaying these. It is time consuming and you have to limit the number ofrelevant variables to make it practical.
22 The three averages and range
There are three different types of average:MODEmost commonMEANsum of valuesnumber of valuesMEDIANmiddle valueThe range is not an average, but tells you how the data is spread out:RANGElargest value – smallest value
23 Comparing sets of dataHere is a summary of Chris and Rob’s performance in the 200 metres over a season. They each ran 10 races.ChrisRobMean24.8 seconds25.0 secondsRange1.4 seconds0.9 secondsWhich of these conclusions are correct?Robert is more reliable.The first and the last statements are correct. The data on the next page will illustrate why the fourth statement is not always correct. The second statement is not correct because a higher mean means he is slower. The third statement is incorrect because a high range means he is inconsistent.Robert is better because his mean is higher.Chris is better because his range is higher.Chris must have run a better time for his quickest race.On average, Chris is faster but he is less consistent.
24 Pie charts A pie chart is a circle divided up into sectors which are
representative of the data.In a pie chart, each category is shown as a fraction of the circle.For example, in a survey half the people asked drove to work, a quarter walked and a quarter went by bus.In a bar chart, the size of each category is compared with each of the others. In a pie chart, each category is compared with the whole.Point out that if the sectors are not labelled we must include a key.
25 How many people are represented by an angle of 36°?
Pie chartsTo convert raw data into angles for n data items:360 ÷ n represents the number of degrees per data item.For example, 40 people take part in a survey. What angle representsone person?360° ÷ 40 =9°two people?9° × 2 =18°eight people?9° × 8 =72°You may want to link this to previous work on ratio and proportion. An alternative method is to use fractions e.g. 1/40 x 360o = 9o and 36/360 =1/10; 1/10 of 40 = 4.How many people are represented by an angle of 36°?There are 9° per person.36° ÷ 9° =4 people.
26 Drawing pie chartsThere are 30 people in the survey and 360º in a full pie chart.Each person is therefore represented by 360º ÷ 30 = 12ºWe can now calculate the angle for each category:NewspaperNo of peopleWorkingAngleThe Guardian8Daily Mirror7The Times3The Sun6Daily Express8 × 12º96º7 × 12º84º3 × 12º36ºTalk through the first method. This method works well if the number of people in the survey (or whatever the pie chart is being used to represent) divides exactly into 360°. Once we know how many degrees represent each person we can multiply this amount by the frequency.Stress that we should check that the angles add up to 360º.(Although, in cases where the angles have been rounded there is the possibility that the angles won’t add up to 360º.)6 × 12º72º6 × 12º72ºTotal30360º
27 Drawing pie chartsOnce the angles have been calculated you can draw the pie chart.Start by drawing a circle using a compass.The Daily ExpressThe GuardianDraw a radius.Measure an angle of 96º from the radius using a protractor and label the sector.72º96º72ºThe Sun84º36ºThe Daily MirrorMeasure an angle of 84º from the the last line you drew and label the sector.The TimesRepeat for each sector until the pie chart is complete.
28 Drawing bar charts When drawing bar chart remember:
Give the bar chart a title.Use equal intervals on the axes.Label both the axes.Leave a gap between each bar.
29 Drawing bar chartsUse the data in the frequency table to complete a bar chart showing the the number of children absent from school from each year group on a particular day.YearNumber of absences7748539321011Start by deciding on a suitable scale for the vertical axis. For example use each division to represent two pupils. Number this axis using the pen tool. Ask volunteers to drag each bar to the required frequency.Copy this slide and modify the table to produce more examples if required.
30 Bar charts for two sets of data
Two or more sets of data can be shown on a bar chart.For example, this bar chart shows favourite subjects for a group of boys and girls.Stress that we must include a key when more than one type of data is displayed in the same chart.Ask,What subject did most girls like the best?What subject did most boys like the best?Is it possible to tell if an equal number of boys and girls took part in the survey?
31 Frequency diagramsFrequency diagrams can be used to display grouped continuous data.For example, this frequency diagram shows the distribution of heights for a group students:FrequencyHeight (cm)5101520253035150155160165170175180185Heights of studentsStress that the difference between this graph and a bar graph is that the bars are touching. Bar graphs can only be used to display qualitative data or discrete numerical data where as histograms are used to show continuous data.Strictly speaking, for a histogram we plot frequency density rather than frequency along the vertical axis. However, this is not make any difference to the graph when the class intervals are equal as in this example.This type of frequency diagram is often called a histogram.
32 Drawing frequency diagrams
Use the data in the frequency table to complete the frequency diagram showing the time pupils spent watching TV on a particular evening:Time spent (hours)Number of people0 ≤ h < 141 ≤ h < 262 ≤ h < 383 ≤ h < 454 ≤ h < 53h ≤ 51Start by discussing how to number the vertical axis. Use the pen tool to do this.Drag the first bar to the appropriate height. Ask a volunteer to drag the next bar and continue until the bars are complete.Finally, ask a volunteer to use the pen tool to show how the horizontal axis should be numbered.Copy this slide and modify the table to produce more examples if required.
33 Histograms and Frequency Polygons
We can show the trend of these graphs more clearly using a FREQUENCY POLYGON.Using a previous example, you first need to draw a histogramThen joint the midpoints of each column.FrequencyHeight (cm)5101520253035140145150155160165170175Heights of Year 8 pupilsPoint out that a frequency diagram is very similar to a bar chart except that the bars touch each other and the divisions between the bars are labelled.Ask pupils to give the modal class interval for the data on the board.Discuss the type of data that would be shown in a frequency diagram. For example, time taken to run a race, foot length, weights etc. In other words, anything that is measured.
34 What does this scatter graph show?
50556065707580852040100120Number of cigarettes smoked in a weekLife expectancyThis data is fictional. However, there is a variety of research linking smoking to a number of fatal diseases such as cancer. For further details, see the ASH website (www.ash.co.uk).It shows that life expectancy decreases as the number of cigarettes smoked increases.This is called a negative correlation.
35 Interpreting scatter graphs
Scatter graphs can show a relationship between two variables.This relationship is called correlation.Correlation is a general trend. Some data items will not fit this trend, as there are often exceptions to a rule. They are called outliers.Scatter graphs can show:positive correlation: as one variable increases, so does the other variableStress that when there in no correlation between the variables it does not necessarily mean that there is no relationship, only that there is no linear relationship between them.negative correlation: as one variable increases, the other variable decreaseszero correlation: no linear relationship between the variables.Correlation can be weak or strong.
36 The line of best fitThe line of best fit is drawn by eye so that there are roughly an equal number of points below and above the line.Look at these examples,510152025Strong positive correlationWeak negative correlation510152025Strong negative correlationWeak positive correlationNotice that the stronger the correlation, the closer the points are to the line.If the gradient is positive, the correlation is positive and if the gradient is negative, then the correlation is also negative.
37 Line of best fitWhen drawing the line of best fit remember the following points,The line does not have to pass through the origin.For an accurate line of best fit, find the mean for each variable. This forms a coordinate, which can be plotted. The line of best fit should pass through this point.The line of best fit can be used to predict one variable from another.It should not be used for predictions outside the range of data used.The equation of the line of best fit can be found using the gradient and intercept.
38 Constructing stem-and-leaf diagrams
The data below represents the numbers of cigarettes smoked in a week by regular smokers in Year 11.Put this data into a stem-and-leaf diagram.The stem should represent ____ and the leaf should represent _____.tensThis data is fictional. Ask questions such as “How many people took part in the survey altogether?” “How many spent under £5/ £7 etc?” “What is the mean/ median/ modal amount spent?”unitsWork out the mode, mean, median and range.
39 Calculations with stem-and-leaf diagrams
ModeThe mode is __ .1 54321Leaf (units)Stem (tens)7MeanThere are ___ people in the survey and they smoke a total of ____ cigarettes a week.22427427 ÷ 22 =___19MedianThe median is halfway between ___ and ___.Ask questions such as “How many people took part in the survey altogether?” “How many spent under £5/ £7 etc?” “What percentage smoke more than 20 a fortnight?” as well as the averages and range.1719This is ___.18
Remember the movie The Social Network?
Yes, it’s about Facebook. Yes, it’s about Mark Zuckerberg, who is currently one of the ten wealthiest people in the world. And yes, it gets ugly at times. Underneath the drama, however, was a very simple but profound insight: Data can change the world.
A better way to put that insight might be: what we do with data can change the world . . . for better or worse! In the case of Facebook, it had some very humble beginnings involving writing an algorithm that used existing data to create an entirely new way of interacting with others in the digital age.
Maybe Facebook changed the world “for better,” maybe it didn’t, or maybe it’s a bit of both. Whatever your opinion happens to be, there’s no escaping the fact that what Mark Zuckerberg did with some calculus chops and some basic data has completely and irreversibly changed the way we interact with each other.
You can study this field in depth with a degree in big data. If you want the flexibility of an online course of study, take a look at our ranking of the Best Online Big Data Degrees.
But that’s just one example. Imagine how much other data is out there about any number of things, and imagine the possibilities of what can be done with it all . . . yes, both good and bad! We are in the infancy stages of even knowing how to collect and store all this data, much less how to interpret it and use it.
Enter the desperate need for technical expertise in this rapidly growing area that has left no industry untouched. Schools are quickly mobilizing to create highly specific programs tailored to the needs we have now for professionals who know how to wrangle terabytes of information, interpret it, and explain to the rest of us how it effects a company’s needs on the ground.
Many schools have responded to this need through Master of Science degrees in a variety of different areas–some in business analytics, some in data mining, and others still in analytics.
They often go by different names, and many have different emphases, but what they all have in common is a commitment to training professionals in big data who can help us all figure out what to do with all the information at our disposal.
What Are the Best Big Data Degrees?
So where are these programs? And where do you start? That’s where we come in. We’ve compiled a list of the top 50 degrees in big data with your needs in mind. Some are online, some not, and some offer flexibility in this particular respect.
The first thing we want to know, like you, is the bottom line: how much is this going to set me back? So we’ve included the annual tuition for each school right up front for you. We’ve also included our College Choice Score, which is computed based on each school’s reputation in the field and its return on investment.
The end result is a list of schools that balance cost, reputation, and, ultimately, what they can do for you and your unique goals.
The rankings you’re about to read are based on a few important sources. The first source is actual college freshman polled during a nationwide survey published by the Higher Education Research Institute at UCLA.
These students rated academic reputation, financial aid offerings, overall cost of school, and the survey also took into account graduate success rates ono the post-college job market. These factors were weighed equally alongside data from other publicly available sources, including U.S. News & World Report, the National Center for Education Statistics, and PayScale.com.
The Social Network was a big hit. Could a Social Network 2 be in the works . . . and could it be about you?
Carnegie Mellon University
College Choice Score: 100.00
Average Tuition: $1,920
Carnegie Mellon has a long history of excellence in technology, engineering, and business education and research. It offers a Master of Science in Information Technology, which can be earned online with a specialization in Business Intelligence & Data Analytics. Working adults and traditional students alike can complete the program at their own pace through Carnegie’s cutting-edge content management tools.
Carnegie’s big data program emphasizes data analytics, management strategy, and information technology. Coursework in the program includes training in economics, business, finances and accounting, geographic systems, database management, statistics, digital privacy, data warehousing, exploring and visualizing data, and applied data mining. Carnegie’s reputation precedes it, and its MS program in Information Technology with a specialization in Business Intelligence and Data Analytics is no exception.
Missouri University of Science and Technology
College Choice Score: 99.40
Average Tuition: $1,230
Missouri University of Science and Technology is one of the most respected STEM-focused public universities in the Midwest. While it is known for its engineering programs, it also offers recognized business programs and online education. Missouri S&T offers an online graduate certificate in Business Analytics and Data Science, which can be applied toward a masters degree in the field for students who want to continue their education.
Areas of concentration in Missouri S&T’s program include business analytics and data science, big data and security, or big data management and analytics. Students will benefit from coursework in the areas of big data analytics, information visualization, business intelligence, cloud computing, machine learning, applied time series analysis, and regression analysis.
Johns Hopkins University
College Choice Score: 99.39
Average Tuition: $1,293
One of the country’s leaders in STEM education, Johns Hopkins University offers students interested in big data two options: an MS in Data Science and a post-master’s certificate in Data Science. Both programs can be pursued completely or in part, which provides options for today’s working professional. Emphases of the program include applied mathematics, statistics, and computer science.
Through its programs, JHU trains students not just to manage data but to interpret and apply it. Its faculty sets a strong precedent for this, being among the leaders in this emerging field. On-site students can also benefit from JHU’s state-of-the-art facilities. Coursework in the program includes the areas of data visualization, computer programming, mathematics, probability, statistics, and simulation.
University of California Berkeley
College Choice Score: 97.87
Average Tuition: $877
Berkeley needs little introduction. It is the flagship institution of the UC system, and it has been one of the country’s leaders in research and innovation for decades. Berkeley is home to the School of Information, one of the first and only centers of its kind. The School of Information offers an online Master of Information and Data Science program through which students have access to the best resources in the field as well as those of the social sciences, computer science, statistics, management, and law.
While Berkeley’s program is offered both on campus and online, online students do have a brief residency requirement as part of the program. The program is designed to train students in the following areas: research design, cleansing, communication, storage and retrieval, statistical analysis, mining and exploring, ethics and privacy, and data visualization. Graduates of Berkeley’s program have obtained employment in a wide variety of industries.
College Choice Score: 97.83
Average Tuition: $1,185
Villanova is regularly recognized for excellence in any number of areas. It is a top-ranked regional university, and it has developed a reputation in graduate programs, business, and online education. Villanova offers a Master of Science in Analytics degree through the Villanova School of Business. The program’s location in the Business School makes it a program with a practical emphasis, which will be appealing to students who are either already in the workforce or anticipate pursuing a career in business.
Villanova’s program is 33 credit hours, and its coursework includes training in data models and structured analysis, multivariate data analysis, enterprise data management, analytic methods for optimization and simulation, business intelligence, advanced applications, and a practicum. The degree can be completed fully online, and its graduates are in high demand.
College Choice Score: 96.98
Average Tuition: $755
Fairfield University, based in Fairfield, Connecticut, is a well-respected regional school in New England. Its graduates enjoy a high starting salary average, making it a good return on investment. It has a growing number of online programs, including data analytics. Fairfield offers a Master of Science in Business Analytics, a 30-credit degree that is no exception to the rule—graduates of the program are in high demand, and starting salaries are comparatively high.
In the MS in Business analytics program, students will take courses in information systems, database management, data mining, business statistics and analytics, and forecasting and predictive analytics. Electives provide additional opportunities to pursue further interests. The program also offers a marketing analytics concentration, which includes coursework in marketing management and research, multivariate data analysis for decision making, experimental research in marketing, and contemporary topics in marketing.
University of Alabama in Huntsville
College Choice Score: 96.90
Average Tuition: $551
University of Alabama in Huntsville has a long tradition of offering technical education in response to demand. Indeed, when a need arose for engineering education for the military, the University of Alabama system opened its Huntsville campus to meet the need. UAH is a major center for technology and research in Alabama and one of the region’s leading schools. UAH offers a Master of Science in Management Science with a specialization in Business Analytics, which can be completed on campus or online.
UAH’s MSMS-BA provides training in advanced analytic tools and techniques to meet the needs of organizations. Students will receive training in detecting patterns and relationships in data, predicting trends, developing forecasts, and improving decision-making within a business environment. They will also learn skills such as data mining, predictive modeling, design of experiments, and statistical analysis. Examples of how these skills can be applied include the ability to better understand customer buying patterns, improve customer retention, assess the impact of marketing programs, manage new product launches, and reduce production, inventory, and shipping costs.
Pennsylvania State University
College Choice Score: 96.88
Average Tuition: $854
Penn State remains one of the world’s undisputed leaders in online education, having been one of the original institutions to establish online offerings nearly twenty years ago. Penn State’s programs in big data draw on the resources of four colleges and departments for the best, broadest, and most thorough education in big data mining, analysis, and communication possible. It offers three programs in analytics: a general Master in Data Analytics, a graduate certificate in business analytics, and a Master of Professional Studies in Business Analytics.
Penn State places an emphasis on data mining, modeling, and architecture as well as extraction, transformation, and loading (ETL) development. Its program can be completed online or residentially, and it is one of the first to offer training from multiple departments and schools. Students can expect to receive training in the areas of applied statistics, predictive analysis, data-driven decision making, database design, regression methods, genetic algorithms, and more.
College Choice Score: 96.10
Average Tuition: $830
Boston University is no stranger to our rankings or those of others. It has world-renowned programs in business, law, and medicine, and it has a growing offering of online programs through its Metropolitan College Online, a high-quality distance education system that provides degree programs for working adults. Through this college, BU offers a Master of Science in Computer Information systems, which focuses on data mining, analysis, and visualization.
BU’s program is shorter than those of some schools, and it also seeks to strike a balance between theory and practice. In addition to its primary objectives of analytics, data analysis and visualization, web analytics and mining, and data mining, the program also offers training in applied probability and statistics, data visualization techniques, and web analytics and metrics. Graduates of the program enjoy an excellent placement record in a wide variety of careers in computer information systems.
Oregon State University
College Choice Score: 95.87
Average Tuition: $576
Oregon State University, a large public institution, has made great strides in recent years with its online offerings. For working professionals in data analytics, those looking to change careers, or students entering the field fresh, OSU offers two graduate options: a Master of Science or a Graduate Certificate in Data Analytics. The programs can be completed online through OSU’s College of Science and Department of Statistics.
Both programs are intended to train students in statistical and analytical skills, and students gain expertise in data analysis, interpretation, and visualization through exploring real-world data problems. The program places an emphasis on quantitative methods, and students can expect to be well prepared for any number of scenarios in statistics, computer science, mathematics, policy, and applied sciences.
Illinois Institute of Technology
College Choice Score: 95.55
Average Tuition: $1,250
Illinois Tech is one of the foremost polytechnic institutes in the Midwest. It offers a Master of Data Science program that can be completed in sixteen months on a full-time basis or two years on a part-time basis. The program emphasizes the proper application of data research, and it prepares data professionals for responsible, ethical design-making through the study of mathematics, statistics, and computer science.
Students will become experts in analyzing data, visualizing results, and articulating discoveries. The program also encourages critical thinking, communication, and good decision-making skills, which are essential to successful job performance. Students will learn to approach data using the scientific method, question premises, think structurally, and find real-world solutions to big data problems.
West Virginia University
College Choice Score: 95.46
Average Tuition: $569
Located in Morgantown, West Virginia, West Virginia University has a long history of providing practical education to the residents of the state. It has recently expanded that mission to online offerings, including a Master of Science in Business Data Analytics through its College of Business and Economics. This degree can be completed in as little as 12 months and requires two brief residencies on campus.
Classwork for this program includes the areas of business intelligence, ethics and data collection, business statistical methods, data mining, simulation modeling, visualization, decision science, management, and a business analytics practicum. Outcomes of the program include facility in statistical techniques, data mining strategies, databases, and sustainable and effective analytic tools.
City University of New York
College Choice Score: 95.33
Average Tuition: $405
The School of Professional Studies at the City University of New York offers ten bachelor’s degrees, eight bachelor’s degrees, and numerous undergraduate and advanced certificates and professional non-degree programs. It is designed to provide online and on campus programs that meet the needs of adults who are looking for a seamless way to finish or transition into a bachelor’s degree, earn a master’s degree or certificate in a specialized field, advance in the workplace, or change careers. The SPS offers an online Master of Science in Data Analytics.
Through CUNY’s program students will build portfolios of increasingly complex projects using popular programming languages such as R and Python, popular languages in IT. Students will also gain experience building predictive and prescriptive models, practice giving presentations, and interactively . The MS in Data Analytics program culminates with a capstone project that represents highly sophisticated, but practical, solutions to address real world problems.
University of Wisconsin
College Choice Score: 94.13
Average Tuition: $675
The University of Wisconsin can be found on rankings lists across the curriculum and in a wide variety of specialty areas. UW features a growing catalogue of online courses as well, and it remains one of the most well-respected public universities in the country in areas as diverse as linguistics, business, and the sciences. UW offers a Master of Science in Data Science. Students are trained in statistical and predictive models for business analytics as well as forecasting strategic methods of management in finances and operations.
UW students are also trained in business intelligence solutions, data warehousing and mining techniques. UW also offers a Graduate Certificate in Business Analytics, which requires fifteen credit hours. Electives in UW’s program include business forecasting methods, web mining and analytics, marketing analytics, database marketing, and business intelligence technologies and solutions.
College Choice Score: 94.10
Average Tuition: $730
Regis University, a private Catholic institution in Denver, Colorado, is regularly featured in rankings of regional universities and online programs. Its College of Computer and Information Sciences offers an online Master of Science in Data Science, a 36-credit graduate degree that brings together technical skills with the Jesuit commitment to social impact, challenging students to consider the implications of big data on business and society. It specializes in the areas of data analytics, data engineering, statistics, predictive analytics, or data visualization.
A unique feature of Regis’s program is that students are required to take courses in the ethics, privacy, and social justice in data science in addition to standard courses in the field. Students in Regis’s program will be prepared to assemble data, identify trends and patterns, implement data to effectively communicate, analyze data to identify new areas of revenue, appraise data for ethical issues, investigate how data science affects society, and more.
Georgia Southern University
College Choice Score: 93.85
Average Tuition: $410
Georgia Southern University, part of the University of Georgia system, has been garnering national attention in recent years for a number of initiatives, perhaps most significantly for achieving status as a national rather than regional institution. It has been acknowledged by many rankings as “best value” school. GSU offers a Master of Science in Computer Science degrees with optional specializations in Data Mining or Data Warehousing.
GSU’s program is a thirty-hour program, and it prepares students primarily for careers in the IT industry. The MSCS seeks to educate students in both theoretical and practical skills, for instance, both data mining and business application. Students can expect to be prepared in areas such as databases, artificial intelligence, Web systems, data mining, data warehousing, and distributed database systems. Areas of practical application include software engineering and wireless and mobile systems.
College Choice Score: 93.80
Average Tuition: $1,500
Texas A&M University, located in College Station, Texas, needs little introduction, particularly in a list of rankings in STEM fields. It is not only one of the largest universities in the country but also one of the leaders in higher education in the sciences and technology. The flagship school of the Texas A&M system, it has also expanded its online offerings in recent years. A&M offers a Master of Science in Analytics through its Department of Statistics.
Coursework for the MSA is offered through the Mays Business School and welcomes working professionals with strong quantitative skills, for example bachelor’s degree holders in the sciences, mathematics, business, and engineering. It is a thirty-six hour degree, making it one of the more demanding programs on our list, but students can expect to be prepared to serve in a number of careers, such as utilities, manufacturing, information, healthcare, natural resources, transportation, warehousing, finance and insurance, and many other sectors.
College Choice Score: 93.70
Average Tuition: $410
The University of Arkansas, located in Fayetteville, Arkansas, is home to the Walton College of Business, an increasingly well-respected fixture on the national management education landscape. Arkansas has also increased its number of online offerings in recent years. Students interested in big data will be interested in two of U of A’s programs: a twelve-hour online graduate certification in Business Analytics, or a Professional Master of Information Systems degree.
Students who complete the graduate certificate can transfer to the MSIS program. Coursework for the program includes training in data management systems, decision support and analytics, and business intelligence and knowledge management, IT ERP fundamentals, e-business development, and/or systems development. An additional benefit of U of A’s program is its partnerships with companies such as IBM, Microsoft, SAP, and others, which offer students opportunities to acquire experience with problems real companies face.
College Choice Score: 93.68
Average Tuition: $895
Does the name “Quinnipiac” sound familiar? If so, you probably heard the school’s name during the 2016 U.S. Presidential race, during which Quinnipiac offered its expertise in data management through a number of well-publicized election polls. Located in Connecticut, Quinnipiac has come into the national spotlight in recent years by investing in online education. Quinnipiac offers a Master of Science in Business Analytics through its online campus, and it is designed with working professionals in mind.
Quinnipiac is dedicated to turning its graduates into assets, and to this end, it offers coursework in a wide range of areas, such as data visualization, mining, management and warehousing, statistical analysis, information technology project management, and predictive modeling. Its graduates are in high demand in the data professions, and they have been placed in careers with marketing agencies, in financial services, insurance, manufacturing, and more.
College Choice Score: 93.45
Average Tuition: $791
The flagship school of the University of Oklahoma system, U of O has a long history of excellence as a large, public institution in areas as diverse as health care, business, and the arts. It is also known for being one of the first public institutions to incorporate digital technology into higher education. U of O offers a Master of Science degree in Data Science and Analytics, an intensive fourteen-month program offered through the School of Computer Science and the School of Industrial and Systems Engineering.
Oklahoma’s program integrates both computer science and industrial and systems engineering. Coursework in the program includes data analysis, data analytics, and systems engineering. Oklahoma’s program is designed to be flexible—it offers both online and on-campus options as well as full- and part-time enrollment, and students can opt for one of two completion tracks: coursework only or research-thesis.
College Choice Score: 93.30
Average Tuition: $850
Auburn University has a storied history both in the South and, more recently, as a national-level research university that has attracted attention for its design, engineering, and business programs. Auburn offers a Master of Science in Information Systems through its Harbert College of Business. It is a thirty-credit degree that places a heavy emphasis on information technology.
Graduates of Auburn’s program can expect to be prepared in the areas of business intelligence and analytics, database analysis, administration, systems architecture, information security, and software engineering. Students will also explore big data issues through courses such as predictive modeling, data communication, quantitative methods, and advanced database administration and development. Students can complete the degree fully online if they wish.
College Choice Score: 93.10
Average Tuition: $1,561
Northwestern University, located in north Chicago, Illinois, needs little introduction. It is a pacesetter not only in national research universities but internationally as well. It leads the pack in any number of disciplines, from business and law to the arts and sciences. Northwestern offers a Master of Science in Predictive Analytics, one of the very first online big data degree programs in the country.
One of the features that sets Northwestern’s program apart from others is the areas of application they prepare students for, such as sports management and performance analytics, web and network data science, advanced modeling techniques, marketing and risk analytics, and data visualization. Students can expect to be prepared for database administration, predictive modeling and quantitative analysis, communication of analytics, and leadership skills. Students can pursue their degree on either a part-time or full-time basis.
College Choice Score: 93.05
Average Tuition: $600
University of Nebraska Lincoln is the flagship school of the University of Nebraska system and one of the leading public universities not only in the Midwest but in the country also. Its online offerings are robust, and students interested in big data will be interested in Nebraska’s Master of Business Administration degree with a concentration in Business Analytics. It will appeal particularly to students who have an interest in applying data solutions to business challenges.
Nebraska also offers a twelve-credit graduate certificate in business analytics, which can be completed in a year or less. Both programs offer training in information technology and statistics, as well as business analytics, econometrics, strategic database marketing, and data mining applications. Whether you are looking for a brief, specialized program in analytics or a more comprehensive program, Nebraska will appeal to a wide variety of students.
Arizona State University
College Choice Score: 93.05
Average Tuition: $1,151
Arizona State is another national-level leader in both on-campus degree programs and online education. It has quickly established itself as one of the standards in online higher education. Its Carey School of Business, one of the largest in the country, in particular has been a pacesetter in distance education for business students. ASU offers a Master of Science in Business Analytics, a sixteen-month program that can be completed on a part-time basis.
Each course in ASU’s program lasts five weeks, and students take classes one at a time. The program requires ten courses total, but students can work at their own pace according to their goals and capabilities. Coursework in the program includes training in enterprise analytics, applied analytics, data mining I and II, data-driven quality management, analytical decision-making tools I and II, business analytics strategy, marketing analytics, and an applied project.
College Choice Score: 92.68
Average Tuition: $745
DePaul University, located in Chicago, Illinois, is a Catholic university with a mandate to provide education access to lower-income and minority students. It has been recognized as the most diverse university in the country by the Princeton Review. DePaul offers a Master of Science in Predictive Analytics through its College of Computing and Digital Media. The MSPA is also affiliated with DePaul’s Data Mining and Predictive Analytics Center.
DePaul also offers a Graduate Analytics Certificate, a five-course program that is less involved compared to the MS degree. Students in both programs can expect exposure to a wide array of subjects relevant to big data, such as programming, tools and techniques for computational analysis, and statistics and data analysis. Additional subjects include mining big data, social network analysis, programming machine learning applications, intelligent information retrieval, and web data mining.
Central Connecticut State University
College Choice Score: 92.38
Average Tuition: $534
Central Connecticut State, an affordable school located in New Britain, Connecticut, has set itself up for success by expanding its online offerings in recent years. It puts excellent higher education in reach for a larger public. CCSU offers a Master of Science in Data Mining, a degree directly relevant to students interested in careers in big data. Coursework can be completed online, but students should be prepared to travel to campus at the completion of their degree to present a final thesis project.