Exlporing your Networks - Assignment 3

Social networks are a peculiar phenomenon to research. However due to the rise of digitization, the internet, globalization and consequently the rise of social networking website, more and more people are able to create ties beyond geographical and social boundaries.

According to Borgatti et al (2009), “One of the most potent ideas in the social sciences is the notion that individuals are embedded in thick webs of social relations and interactions.” This idea can be seen in real life in networking.  One of the most prominent online examples is Linkedin. For example my own Linkedin Network shows  the people with whom I share a social relation and with whom I interact. Each cluster is based on a different shared factor, which is indicated by different colours.

At the beginning of 2014 Linkedin launched a new service based on the social relations and interactions, and more specifically the theory of ‘six degrees of separation’ which is based on the small world problem by de Sola Pool  & Kochen (1978).  The six degrees theory says that “because we are all linked by chains of acquaintance, you are just six introductions away from any other person on the planet” (Smith, 2008).  Linkedin’s feature allows users to see how exactly they are connected to another person as well as the degree by which they are connected, helping to visualize the thick web of social interactions and the practical results thereof.

However, how can you yourself visualize your linkedin network? Well that can be done quite easily through a programme called Socilab. This is a website which will create a nodes cloud to show all of your connections (up to 500 of them). This offers a fast look at overall connections.  In Socilab you can choose to colour the nodes either by industry, country or location.

One way to analyse your output is through the ‘ego networks’ method (Hanneman & Riddle, 2005). This method suggest that every node can be considered as an ‘ego’. The ego can be any type of entity as long as it has a node in the network.  If you take a single step in any direction from a node, you encounter what Hanneman and Riddle refer to as a "N-step neighborhood".

By knowing your position and the position of others in the overall network you can see who In your network has a brokerage role between yourself and another ego, or even how you yourself acts as a broker between nodes. The possible roles one can take on in brokerage are (1) coordinator, (2) consultant, (3) gatekeeper, (4) representative and (5) liaison. By knowing where you stand in a network helps you to create understanding of the world around you and your role in the world both online as well as offline.

So now that you know a bit more about networks, maybe it’s time you visualized your own network and how it can help you further.

P.S. Because of the structure of the table generated by Socilab it is not effective to create a clear visualization on Tableau as each person is labled as a dimension as well as a measure.

Borgatti, S. P., Mehra, A., Brass, D. J., & Labianca, G. (2009). Network analysis in the social
                sciences. Science, 323(5916), 892-895.

de Sola Pool, I. and Kochen, M. (1978). Contacts and influence. Social Networks, 1(1), pp.5-51.

Hanneman & Riddle (2005). Introduction to social network methods- Chapter 9, Ego Networks.
                Online http://faculty.ucr.edu/~hanneman/nettext/C9_Ego_networks.html

Smith, D. (2008). Microsoft proves there are just six degrees of separation between us. [online]
                the Guardian. Available at:
                http://www.theguardian.com/technology/2008/aug/03/internet.email [Accessed 28
                Nov. 2015].

Textual Analysis & Tableau for Twitter | Assignment 2

By Aeyiondy Dorant & Xaviera Huisman

For Social Sciences qualitative research methods are a valuable tool to really understand how reality is socially constructed and to grasp the meanings that are created by people, be it individuals or groups. As a result, text analysis is seen as one of the most important methods that can be used in cultural studies. “When we perform textual analysis on a text, we make an educated guess at some of the most likely interpretations that might be made of that text” (McKee, 2001, p. 3). While conducting textual analysis it is not prefered to find the correct interpretation of the text. According to McKee (2001) “there is no such thing as a single, ‘correct’ interpretation of any text. There are large numbers of possible interpretations, some of which will be more likely than others in particular circumstances” (p. 4). With textual analysis it is of even higher importance to understand how we make sense and share this created sense with the rest of the world. This is imperative as there are an array of different meanings attributed to a certain topic by people. All in all it is important to conduct a text analysis in order to get a grasp on what kind of meaning is given to the world. With that in mind we would like to elaborate on the fact why it is considerable to conduct textual analysis on the following case.

On November 21st, a week after the Paris attacks, Brussels was reported to be on lockdown as a result of ‘imminent’ terror attack threats (Murray et al., 2015). Using the website www.hashtagtracking.com, and through browsing Twitter we noticed that three English hashtags were prominent in tweets regarding the lockdown namely, #brussels #brusselsalert and #brusselslockdown.

The types of Tweets utilizing these hashtags represent what Menchen-Trevino (2013) refers to as participation traces in the field of digital behaviour. Menchen-Trevino sees participation traces as a vital part of social scientific research “Participation data forms a very small part of overall digital trace data, but it is particularly important in social science research, as it is often a record of intentional actions” (p. 330). Tweets although part of participation data, cannot be considered as semantically rich, as they only allow the use of 140 characters. Yet even with the limited semantics people are able to convey meanings through tweets in real time, as a situation is occurring. Not only that, but tweets also show the social aspect of meaning creation, as others can react and add their insights by retweeting, replying and liking. For this reason we believe that tweets, both text and image based, are perfectly suited for textual analysis or semiotic analysis.  

Take for example the request by officials to not tweet about ongoing Lockdown, has lead to the collective recreation of meaning by posting the pictures of cats with the hashtag instead. Tweets like these show us how people are making sense of the world and the situation they live in (McKee, 2001).

So how then can these tweets be analyzed for meaning?

By using the Twitter Web application for tableau it is fairly easy to create a text based dataset surrounding a relevant hashtag. By inserting these texts into special text analysis software such as Context and Sentistrength, one can determine the themes and topics present in the tweets, as well as determine the most prominent sentiments.
To take it a step further and to visualize the textual analysis, the Tableau software offers a handy set of tools with which one can actually visually show which themes are most recurrent and which sentiments are mostly present in the social chatter on twitter. Furthermore Tableau can show the different regions that are participating in the creation of meaning.

This was this weeks exploration of Tableau and other Digital Research methods.

SOURCES

McKee, A. (2001). A beginner’s guide to textual analysis. Metro Magazine, 1 - 21.

Menchen-Trevino, E. (2013). Collecting Vertical Trace Data: Big Possibilities and Big
Challenges for Multi-method Research. Policy & Internet, 5(3), 328{339.
doi:10.1002/1944-2866.POI336

Murray, W., Gayle, D., Fishwick, C., Traynor, I. and Shaheen, K. (2015). Threat of

'imminent' Isis attack puts Brussels on lockdown - live. [online] the Guardian.

Available at: http://www.theguardian.com/world/live/2015/nov/21/paris-attacks-three  -suspects-including-alleged-scout-arrested-in-turkey-live [Accessed 21 Nov.
2015].

My first Encounter | Assignment 1

During my first Digital Research Methods class I encountered a new programme, which I had never heard of before. Something called Tableau. It looked pretty cool and I could already envision using it for my thesis to create some amazing visuals and tables. However, the whole programme looked to me like some high-tech futuristic NASA certified software, and I wondered how in the world I would ever grasp it. So I decided to create this blog, documenting my journey in discovering Tableau. starting with this post I will be posting my explorations through each of the assignments I have to make for class, and hopefully at the end of this term I can say that I understand at least a little bit of this daunting software.

Assignment 1

For this assignment, I decided to look at the Global Superstore 5Kdataset.

Step 1 

Firstly, I wanted to know how people choose to have their products shipped in different markets and different market segments, based on how much discount they receive.

Dimensions are the categories by which we can divide the data for analysis. For example in the Superstore dataset I decided by using a tutorial to look at the shipping mode of products per  market segments (which I put in rows), and analyze these two dimensions by comparing them for different country markets and BY the quantity of products sold, these two measures were put in the columns section.  

 (Tableau Sheet 1) 

From the visualization I deducted the following:

1. For each customer segment, Standard shipping is the most preferred method of shipping, especially when the discount amount is higher. The consumer segment, which sees the highest amount of discounts, also has the most sales and the highest amount of standard shipped products.

2. There is a huge market gap in Canada, as there are no shipments at all for the corporate and home office segments, and only a very small amount of shipments for the consumer segments, I deduct that Canada is an uncharted market which can offer opportunities for the company.

3. Sales in the EMEA market is behind on the other markets, which can mean 2 things:

- The market recently opened and the company is still gaining foot - The company is not doing well in this market and should look into reasons as to why this could be happening.

I figured out that by analyzing one thing, can lead to the discovery of other facts, which is quite interesting.

Step 2 

This discovery about the markets led me to want to research more about the lack of sales in both Canada as EMEA. Thus I opened a new sheet and did the following:

1. I added Market and order date to the columns

2. I added Segments and Sum of sales to the rows. This is because I still wanted to see how the sales performances were per each customer segment. 

(Tableau Sheet 2)

From the results I can see that data for all markets go back to 2011. When looking more closely at the 2 markets that are behind on performance I saw that EMEA has been improving over the years in all three customer segments. Additionally, the consumer segment is the biggest segment with approximately 200k total sales in 2014. Somehow in 2013 the sales in the consumer and Home office segments underwent a slight increase in growth rate, whilst the corporate segment maintained a steady linear growth. Indicating that whatever factor caused the growth increase in the Consumer and Home office markets did not influence the corporate market.  Additionally when comparing it to the other markets, we see that the factor which brought on this change impacted a few other markets as well. EU, LATAM and US markets all saw a similar increase in sales growth in the same period. Interestingly enough, Africa saw a mix of sales growth increase as well as decrease in the similar period, varying per customer segment.  

Step 3 

The change in growth rates leads me to question what factor could have led to this. I wondered if there was a relationship between the discount and the sales.

To analyze this I put the Market and the Year in the columns, the Segments and the sum of all discounts in the rows section. Additionally I put the sum of all products sold (quantity) as a label so I could compare the height of the discounts with that of the quantity of products sold.  

(See Tableau Sheet 3)

From this table I see that between 2013 and 2014, the exact period in which most markets saw a growth rate increase, we also see an increase in discounts offered which lead to an increase in total products sold. Ultimately, looking back at Canada which is the market that performs the worst, I see that the company is not offering any discounts unlike in the other markets. This could be the reason why there are so little sales in Canada.

Final Thoughts 

This assignment got me acquainted with the Tableau programme and showed me that it is quite a useful programme to use especially for my thesis later on. However, I did need the help of a YouTube tutorial to understand what I was doing and what I was looking at. I have to say that because of the many functionalities that this software has it is quite difficult to get acquainted with and to master. Follow me through my discovery of this software over the next 9 weeks. For more information on Tableau and an easy to follow tutorial click here.

X,

Dee