Correlation or Causation: The most common errors made in data analysis

The research field is prone to errors with significant consequences on generalizing results and interpreting statistical relationships.The following are some of the most common:

statistical errors

  1. Assuming you have a representative sample: This error is common especially in large data sets (big data). The results may look convincing, but the sample does not represent a true picture of the target population.
  2. Measurement errors: Assuming the data has been measured correctly.  Often we get more bad data than good data and you need to adjust for this.
  3. Correlation versus Causation  – There is no statistical test to prove/ disprove whether two things are correlated versus whether one is causing the other to occur. Yet many researchers make this common error of assuming causality instead of correlation.
  4. Assuming that the data represents what you think it does: For example, not because a large proportion of prisoners are from high school A, means it’s correct to conclude that high school A is responsible for producing criminals.
  5. Reliance on features that are destined to change. Some numbers are bound to be way off when circumstances change. Daily and weekly fluctuations are easy to notice, but predicting what will happen one year from now is a different story. No matter how well a research explains current and past results, predicting the future based on current information has many limitations.
  6. Not understanding the impact that algorithm (different statistical tests) choice have on final results. Using different types of algorithms — say, random forest instead of logistic regression — may often have a significant impact on final results.

Can you think of any common errors you might want to add? Please share in the comments below

There is nothing Boring About Statistics

“There’s nothing boring about stats” – Hans Rosling. In this one hour documentary, Hans Rosling highlights the joys of statistics and the inherent beauty in the numbers.

Research Purpose in the Information Age

Understanding Dependent & Independent Variables #Animation

In doing research and data analysis, many have struggled with differentiating between the dependent and independent variables when running statistical tests and/or experiments. A clear understanding of these two concepts will play a major role in an individual’s thinking as a researcher/ experimenter.

This animation seeks to address this common challenge of differentiating between the independent & dependent variables in a research.

Are You Satisfied?

What is satisfaction? How do we ask questions about satisfaction? Is there a best approach?

Love customers

Here we present some tips and best practices for asking satisfaction questions in a survey.

At Balcostics our mission is to empower leaders with the required data and information to make better decisions. Learn more about our full list of research outsourcing services for individuals and companies: Click here

Qualitative & Quantitative Sampling Methods

Research is a systematic approach to collecting, entering and analyzing data. The process of data collection begins with determining which ‘sampling method(s)’ one will use – qualitative or quantitative.

img_20161107_130037

This slideshow provides a practical ‘how-to guide’ in sampling techniques, looking at the qualitative versus the quantitative approaches.

 

At Balcostics our mission is to empower leaders with the required data and information to make better decisions. Learn more about our full list of research outsourcing services for individuals and companies: Click here

The Art of Survey Design

Surveys like products work best when we have a well thought out design. More researchers need to pay attention to how they design their surveys. We found this insightful slideshow which highlight some tips and best practices.

 

At Balcostics our mission is to empower leaders with the required data and information to make better decisions. Learn more about our full list of research outsourcing services for individuals and companies: Click here

Market Research & Emerging Markets

Another thoughtful answer to a question many have asked: How do we approach market research in an emerging market like Jamaica?

 

IMG_00000153We found this insightful, yet simple answer on Quora:

“Market research is a very important part of a business and when you are entering into an emerging market it is of utmost importance that you have put enough efforts in the market research.

You can avail the secondary data through internet, periodicals, research papers, group literature, trade associations, etc. All of these data you can avail at a very minimal expenditure.

For the primary data you really need to spend the money. Money spent on consumer market research depend upon type of customer segment you want to deal with, where your customers are located and type of business (B2B or B2C).

For example if your customer segment comprises of older people it is evident that you can not float a survey to collect data on internet as they are not very familiar with filling surveys online; so you have to go out and complete the survey manually. Hence, that will increase your expenditure. Similarly, if you are dealing in a B2B business money and efforts spend in market research will be less as compared to a B2C business.

Also, just don’t put too much efforts and money in market research, you should have adequate data with you and with that data enter into market. Don’t just miss your time in doing more and more research (I know that’s too bookish but it’s true).”

See the original question and answer here

Market Research Basics

The following question was asked on Quora: “How do you conduct market research?”

This answer by Cate Riegner, was as simple and concise as we could do, so here is it:

Balcostics Research

“Think in terms of different methodologies and consider which ones are right for your needs at various stages of business development, product development and marketing:

1. Syndicated/secondary research: This would be research conducted on/for a particular market that provides market sizing information, penetration data, trending data, etc. Search for published findings at trade organization sites, investment sites, etc. The challenge is finding a study that is relevant to your specific needs, and many charge a fee for a published report.

2. Custom quantitative: If you want quantitative data or feedback on a very specific subject related to your product (e.g., behaviors, attitudes, product concept testing), consider doing your own survey.  The challenge here is making sure your sample is “representative” of the market you want to study, your survey is designed well (e.g., no leading questions, appropriate scales), you process and report data accurately. There are great, affordable survey programming/hosting tools out there (e.g, SurveyMonkey), but you need to pay for sample, survey design and possibly analysis depending on your needs.

3. Custom qualitative: This might be user testing, ethnography, interviews, blind-shopper stuides, focus groups (either online or offline). People put a lot more weight on quantitative data (because it is statistically reliable), but I am a huge advocate of qualitative research for deeper, richer insights.  Don’t discount it just because the sample size is small. The question you need to consider here is: Which approach/method is best for what I want to learn/problem I want to solve?  Each of the types of qualitative research I listed above is better for different needs.

4. Analytic data: A lot of data is generated for free online, through traffic logs, blogs, etc. The challenge here is collecting the data and conducting an analysis/generating findings that can be applied to your specific question/problem you are trying to solve.”

Check out all the answers to the above question on Quora.

SMEs & Market Research

Many Jamaican entrepreneurs (or entrepreneurs to be)  have asked/ or is still asking themselves’ the above question – is market research needed for my new idea?

We’ve found some external insights on the topic, which we hope will provide additional clarity.

 

Galleon (9)

Question as answered by Amy Shropshire on Quora:

Most of the mistakes I’ve seen businesses of any size make have been due to the thinking they know the market place rather than taking the time to actually find out.

For the past 15 years, I’ve been advising clients in the nonprofit and start up world. What I love about them is their eagerness to get going, get their product or service into the hands of their market, and see their dreams come true. The challenge comes when I ask them questions about their market and they can’t answer them. They know all about them, and nothing about their consumers. They expect to put their product out there, do “marketing,” and make millions.

I eventually created a workshop called “No One Cares About You.” where I explained that deep down, consumers don’t really care about you. They care about how you’re going to solve their problem/need/want/desire/etc. And that’s what market research tells you how to do.

I disagree with what Shashank PS says about a completely new product being unlikely benefited from market research. This is where research will benefit you the most – Who exactly is your market? Does the market even want your product/service? Does it fill a need they have? If they don’t know they need it, how can you convince them they do? What channels do they like to be reached on. How much would they be willing to pay? The questions are endless.

I find that most people want to find a way to skip the market research part because its seen as boring. They want to jump right into the fun stuff! But let me finish by saying that the better you know every facet of your market, the more they will want to know you. =)

Question as answered by Matt Bertuzzi on Quora:

There are compelling arguments here for both sides. I have 2 quick thoughts:

1) Market research can hugely beneficial – not unlike learning about the structure of the face and effects of physics. Market interactions are like getting punched in the mouth 1st hand. Imho, you learn more from the 1st and know more from the 2nd.

2) I had no idea I wanted 1k songs in my pocket until I heard Steve Jobs tell me I could have it. Similarly, I might have scoffed at the idea/concept of Twitter in a conference room. But after week 1 using it, I was hooked.

Just my $.02.

See more interesting answers to this question here

At Balcostics our mission is to empower leaders with the required data and information to make better decisions. Learn more about our full list of research outsourcing services for individuals and companies: Click here

The Bigness of Big Data: Video

In this Hand-animated video, the presenter examines the following information on big data:

  1. The connection between big data and the cloud
  2. Classification of big data – what really make ‘big data’ big?
  3. Uses of big data
  4. The value of ‘big data’

Big Data Explained: Video

The following questions relating to big data are addressed in this short animated video:

  1. What is big data?
  2. How does ‘big data’ work?
  3. Where does big data come from?
  4. What are the three ‘V’s of big data?
  5. Who are the users of big data?

Another look at Questionnaire Design

“It is not every question that deserves an answer.” – Pubius Syrus

business decisions

Questionnaire design is  imperative to the success of a research project. When all is said and done, the results of your research reflect the questions that were asked. This slide will highlight the following:

  1. What questions should be asked
  2. Best approaches in phrasing a question
  3. Arranging the questionnaire
  4. Pretesting ones questionnaire

Best Approach To Online Research

True or False? The majority of us will start the research process online – searching for relevant literature relating to the specific research problem we are looking to solve. The following infographic highlight different resources and best approach in doing online research. We believe you will find it helpful

Research is Not Data Gathering!

Here we seek to highlight answers to the following questions:

  1. What is research?
  2. What is the research process?
  3. Why we do research?
  4. Why do a literature review?

Planning your research: The simple yet effective whiteboard

Before embarking on a new research project, the following are considered extremely important to the successful execution of same:

Brainstorming:

Planning should begin with the brainstorming phase, where the problems/ questions that need answering is verbally aired, reviewed and discussed. A white board is an excellent tool that can be used to capture the information and foster the discussion surrounding the problem. Generally most of your research questions come from these brainstorming sessions.

In my experience with Balcostics and helping people with their analysis, I’ve come across many persons who went and collect data that was not relevant in answering/ solving the problem they commissioned the research to solve. Dwayne Spradlin highlighted in his Havard business review article “Are you solving the right problems?”, the importance of asking the right questions and thoroughly defining the problem.

Organizing:

Organization is key! Most successful researchers have good project management skills, whether they were formally trained or not. Tasks listings, resource allocation, budget and timelines are necessary evils in the planning process.

Research is a methodology, Not percentages!

Research is a systematic way of collecting, analyzing and presenting data. The methodology used in the design, collection and analysis of the data is imperative. The methodology should be given high priority, to ensure the sample is representative and margin of errors are adequate.

Please comment and share your opinion on any of the simple points presented.

Visionary Entrepreneur

About the author:

Luwayne Thomas Bsc., Msc.

Co-founder & COO @balcostics

Follow on twitter: @LuwayneThomas

 

Is there a methodology for collecting and analyzing big data?

Jahangir Mohammed provided the most detailed response to our question on Quora – September 2, 2012. We wanted to find out if there was a systematic process/ method involved in the analysis, collection and presentation of big data. Here is Jahangir’s answer:

“I am inclined to say the approach is definitely systematic, but there are lots of options and one needs to figure out what is the best implementation for their specific use case.

 Data collection:

There are various distributed data collection and aggregation frameworks like Flume[1], Chukwa[2] and Scribe[3] which can be leveraged efficiently to collect and aggregrate data in real-time from lots of servers.

If one has the data in some form sitting in RDBMS, they can use sqoop[4] to transfer data between RDBMS and to a big-data framework like Hadoop[5](meant HDFS).

Data analysis:

Hadoop[5] is a well-known framework that allows distributed processing and analysis of big data. There are couple of other frameworks like Cascalog[6], storm[7] – stream processing, some MPI frameworks and some BSP frameworks(like Apache Hama[8]) and Dremel’s open source (is currently being worked on) all of which are created to crunch big data. Also, there is Amazon’s EMR[9] or Google’s big query[10] from a cloud perspective, but to keep it explicit there is nothing stopping to run any open source
implementations on cloud.

Presentation/data visualization:

This can be home-grown to using a commercial product. Some of the offerings out there like Datameer[11] and big query[10] do offer some visualizations, dashboards, excel capabilities and so forth.”

[1]. http://www.cloudera.com/blog/201…

[2]. http://incubator.apache.org/chukwa/
[3]. https://github.com/facebook/scribe
[4]. http://sqoop.apache.org/
[5]. http://hadoop.apache.org/
[6]. https://github.com/nathanmarz/ca…
[7]. https://github.com/nathanmarz/storm
[8]. http://hama.apache.org/
[9]. http://aws.amazon.com/elasticmap…
[10]. https://developers.google.com/bi…
[11]. http://www.datameer.com/

Feel free to leave a comment and add your views in the comments section.

Special thanks to Jahangir Mohammed and Vijay Kamath who both took time out to provide answers to our question.

An Analysis of Violent & Property Crime in Jamaica

This study sought to investigate the impact of police population, arrests rates, unemployment, educational attainment, average income, income inequality and degree of garrisonization on the incidence of violent crimes and property crimes in Jamaica.

Main Findings:

The rate of arrest and average income variables were found to be robust (strong) determinants of violent crime, with both exhibiting a negative relationship. That is, an increase in either arrest rates or average income would be associated with a decrease in violent crimes.

The results for property crime were quite different; the arrest rate was not found to be significant, while average income was found to be positively correlated to the incidence of property crimes. Income inequality (as measured by the Gini Index) also proved to be a positive determinant of property crime.

The policy implications of this paper are as follows:

1. In order to combat violent crime local authorities should focus on improving the arrest rates for violent offenses as opposed to trying to influence any of the other variables investigated in this paper.

2. Though it may be tempting to tout economic growth as a strategy for reducing violent crimes, this would result in a trade-off between violent crime and property crimes, with the latter responding far more sensitively.

3. Reducing the inequality gap is the most effective means by which the authorities may pursue a reduction in property crime. Although average income was also found to be a significant factor, it exhibited a positive relationship, meaning that economic growth should not be expected to deter property crime. On the contrary, the results suggest that property crime in Jamaica is likely to be exacerbated by growth, unless the growth effect is offset by an aggressive reduction in income inequality.

Author: Nadia Grant- Reid, Bsc., MSc.
UWI Lecturer of Economics (Western Campus)

At Balcostics our mission is to empower leaders with the required data and information to make better decisions. Learn more about our full list of research outsourcing services for individuals and companies: Click here

Conducting Online Surveys: Pros & Cons

With the explosion of internet usage globally, it was inevitable that surveys would eventually utilize the online medium to capture required data. The advent of Facebook and Twitter, has further resulted in a paradigm shift – it is now easier to find people on Facebook/ twitter than at a park (base on our informal observations). With smartphones as well, people now have access to all their social media and email messages instantly. It is because of this, why many researchers have opted for the online route to garner and sometimes cajole responses from online users. In the process of collecting data online however, one should be aware of the following:

Some Advantages of online surveys:

1. You can do it yourself! Yes you can! It is really really simple.

2. It cost significantly less – doesn’t require large sums to execute your data collection and entry

3. Diverse distribution channels – survey can be accessed via email, facebook, twitter and on a website. You are allowed to question the people you wouldn’t meet in the park!

4. Convenience – respondents (people completing the questionnaires) have autonomy over time and place where they choose to answer survey questions.

Some Dis-advantages of online surveys:

1. Low response rate – people have friends to talk to, pictures to see, who has time to answer questions?

Possible Solution: People respond to incentives, so ensure as an order of priority that there is a ‘good enough carrot to entice the rabbit’. There needs to be an opportunity to win something special e.g Iphone or a gift certificate

2. Sample not representative – this is especially true when you’re distributing a survey on facebook/ twitter. You have no control over who is answering. You might want only Jamaicans, or persons within a specified age group who are Jamaicans! The fact remains that people outside your specified demography can and will answer your survey – especially when there is a good prize to be won.

Possible Solution: only distribute general surveys on Facebook/ Twitter, where the demographics of the sample is of little significance.

3. Missed deadlines – since people have complete autonomy as to when and where to answer your survey, the responses might be slow in coming, which might delay your entire project.

Possible Solution: In the case where you have an email listing – follow up emails and possibly phone calls would be appropriate to remind persons to complete survey.
In the case of twitter/ facebook – it would boil down to an incentive issue again! Where early respondents would be better place to win, or something special (similar to ladies free before 12pm).

Please leave a suggestion/comment, we do appreciate it.

 

At Balcostics our mission is to empower leaders with the required data and information to make better decisions. Learn more about our full list of research outsourcing services for individuals and companies: Click here

Conducting your own survey? Here is a quick guide

What is a survey?

It is  the systematic/structured gathering of information from a sample/ proportion of individuals for the main purpose of making conclusions/predictions about a larger population/group.

Prior to starting a survey, one should…

  • Define a clear purpose for the survey
  • Create research objectives/research questions you intend to answer
  • Decide on a methodology to collect the data required

 Popular ways of administering a survey:

  1. Online via a website
  2. Online via email campaign
  3. Face-to-face interviews
  4. Interviews via telephone

In general, the method you choose to administer your survey questionnaire will be dependent on the following: Read more of this post

The Telephone Interview Questionnaire

Be Prepared to Interview

Be prepared, Be very prepared. It is one thing to gather the courage to walk up to a perfect stranger, put on your best smile and ask them to fill out a questionnaire. It is a completely different thing to make a  phone call to a complete stranger and ask them to sit on the phone with you for God knows how long, while you ask them questions they don’t want to hear, and can’t be bothered with. Seems harsh but it’s true. The following are some of the main reasons you should be well prepared: Read more of this post

Questions for the Questionnaire

The following represents questions one should ask when creating a questionnaire.

What does the question contribute?

Each question included in the questionnaire, should contribute significantly to acquiring relevant information for the researcher. The questionnaire should be assessed prior, so as to ensure that all the questions present are relevant and if not, irrelevant questions removed. Read more of this post

Questionnaire Design most important

Over the years doing data entry, data analysis and interpretation of results for students, lecturers and companies; We have come to concluded that questionnaire design is one of the most critical element of any survey type study (studies where questionnaires are used to collect data). The questionnaire can make or break your data analysis, so pay attention!

We have seen the good, the bad and the indifferent, as it relates to questionnaires. The good ones always include: Obvious headings, clear instructions as to what should be done to complete the questionnaire, concise questions, and most importantly, the questions asked, should adequately address the hypotheses being tested – the main issues or problems being examined. Read more of this post