Step 5: Using Netdraw
5.3 Changing the colour of the nodes
Properties → Nodes→ Color → Attribute Based (Figure 18).
Figure 18: Change the colour of the nodes
This will open the following window (Figure 19) with a list of the attributes defined in the database.
Figure 19: Select attributes
To change the colour of the nodes, select one of the attributes. “Type of actor” was chosen for this example. A window will appear with various colour options (Figure 20) for the different attributes assigned to the actors. Assigning different colours to the nodes makes it easier to identify tendencies from the map. The colours for the different groups can also be changed by double clicking on the type of actor in the
´Color Nodes by Attribute´ window.
Figure 20: Define the colour of the nodes
5.4 Colouring the lines
There is a way of changing the colour of the lines (Figure 21) to differentiate types of relationships (in this case, information).
The step is: Properties → Lines → Color → Tie Strength.
Figure 21: Change the colour of the lines
In the window that appears (Figure 22) choose the information (values) whose colour is to be changed.
Figure 22: Choose which tie to colour
Below is a chart showing the social relationships in the network (Figure 23).
Figure 23: Network with coloured ties and nodes
The thickness of the ties can also be increased but only if they are encoded between 1 and 5 as shown in the second survey above.
Go to Properties→ Lines→ Size → Tie Strength (Figure 24).
Figure 24: Show the strength of the ties
5.5 A first analysis of the network
At first sight the network looks very complicated but it is possible to focus on specific groups of actors so that the analysis becomes simpler and tendencies are easier to see.
Go to Nodes in the window on the right-hand side of the screen (Figure 25) where the different agrupations of the actors according to their attributes. For this example we chose “Producers”.
Figure 25 Choose groups of nodes
If groups of actors are unselected in this box, they will temporarily disappear from the map, as seen below with the producers (Figure 26). Using this tool the relationships within and between different groups can be analysed either in isolation or overlapping with other groups depending on the purpose of the analysis.
Figure 26: The network without one group of actors
To bring back the invisible nodes, either tick the empty boxes in the window to the right of the screen or on the ~Del button on the tool bar.
This same procedure can also be repeated with the ties. The steps are similar: In the window on the right where it says relations, tick the type of information that is to be focussed upon (Figure 27).
Figure 27: Choose ties
Then in the next window choose the value of the relationship which is to be shown according to its code (Figure 28).
The following options are available:
>
≥
=
Figure 28: Show individual ties ≤When one or more groups of ties are eliminated then some nodes will become isolated from the rest of the network cluster. These can be left in the picture or hidden. To hide isolated nodes go to the tool bar and click on the Iso button (Figure 29).
Figure 29: Hide the isolated nodes
The structure of the network will have changed, creating loose groups separated from the main network cluster meaning that this specific subject of information does not reach the entire network and sub-clusters may have formed with no direct relationship between them.
To return to the complete network map ties click again on the empty boxes in the window, making sure that all ties are visible in the box in lower right hand corner (>0) or on the ~Del button on the tool bar.
This is a basic introduction to the analytical possibilities of this tool but there are many other features. It is important to try out the other options offered by the software.
5.6 Saving charts and files
It is now time to save the work done (Figure 30).
To save go to File → Save Diagram as and choose the file type.
From this programme the maps can be saved in different formats such as jpeg, bitmap or metafile.
Figure 30: Save the chart
In order to save the data that has been worked on so that the next time the file is opened the nodes are the same colour and in the same position, choose the Vna format.
Go to File → Save Data as → Vna → Complete (Figure 31).
Figure 31: Save the file
Step 6: Try for yourself
This manual is a very basic introduction to how to use this software. It is hoped that it will provide a sufficient understanding to give the user confidence to start to experiment with the tool and discover potential applications for their own work.
5. The 2-mode network
The 2-mode network is a more complex way of showing information as it has two different types of nodes, for example to visualize the relationships among actors and their information requests in order to facilitate analysis of often very complex relationships between different social groups and certain subjects.
Below is an exercise which will demonstrate how to use the 2-mode network. 2 modes will be used: the actors themselves and their different information demands to see how they are distributed in the map. Please note the data used in this exercise are simulated.
The Ucinet©∗ programme is needed to create a 2-mode network. This programme is not free and requires a license. It can be downloaded from the internet at the following address: http://www.analytictech.com/ucinet.htm
Step 1 Survey structure
The same considerations as those above apply to preparing a survey to collect information to create a 2-mode network.
The survey structure is a simple table of two columns as seen in table 3. There will be a list of information each actor would like to receive (i.e. the information that they need) and the communication media by which they would like to receive it. The variables suggested in this table can be changed to suit the specific needs of each study.
Table 3: Structure of 2 mode survey
Information demand Communication media by which they would like to
receive it
∗ Borgatti, S.P. , M.G. Everett & L.C. Freeman. 2002. Ucinet 6 for Windows: Software for Social
Step 2: Preparing the database
This produces a three-column database (Figure 32). The first column represents the person interviewed, repeated for each separate demand, the second is the list of the information demands and the third is the means by which the producers wish to receive the information.
It is possible to create a 2 mode network with only the first two columns.
Figure 32: 2-mode network database
This data is then used to create a table. As before all data has to be codified. In this example the desired communication media for receiving information is the tie variable.
Go to Tools → Sort (Figure 33).
Sorting the data in the 3rd variable column makes encoding much easier.
Figure 33: Organise the database
Substitute each different response with a numerical value to make a simpler list.
Remember to make note of which number represents which value.
Figure 34: Encode the tie information for a 2-mode network
Once the information is encoded, it needs to be placed in a matrix (Figure 35). The list of actors is the first column and the list of information demands is the first top row.
The easiest way of doing this is to organize the information demands using the Sort function in excel and one by one fill in the columns of the matrix with codes representing the different communication media, which have already been assigned a numerical value (although if only using the first two columns these will be binary).
Figure 35: Transfer the information from the columns to a table
Once all of the data have been entered the result is a matrix as shown in Figure 36.
Figure 36: A 2-mode table in excel
Step 3: Transferring the data to Ucinet
©The next step is to open Ucinet©. When it opens the screen will be blank.
Go to Data → Spreadsheets → Matrix (Figure 37).
Figure 37: Open a table in Ucinet
Once the Ucinet spreadsheet is open, copy the excel table and paste it into the Ucinet table (Figure 38).
Figure 38: Copy the excel table to the Ucinet table
It is important to check that the dimensions of the matrix as shown to the right of the spreadsheet (Rows / Cols) agree with the values of the matrix that has been imported (Figure 39).
Figure 39: The Ucinet table
All the empty spaces must be filled in with an 0.
To do this go to Fill → Blanks w/0s (Figure 40).
Figure 40: Fill in the empty cells
Save the file as a Ucinet database Go to File - Save as (Figure 41).
Note that all Ucinet files are saved as (.##h).
Figure 41: Save the Ucinet file Now the data is ready to import into Netdraw in order to visualise the network.
Step 4: Visualising 2 mode networks with Netdraw
Using Netdraw to create a 2-mode network is very similar to creating a 1-mode network except that now the data is in a different format.
First open Netdraw,
Go to File → Open → Ucinet Data Set
→ 2 Mode Network (Figure 42).
Figure 42: Open the Ucinet file in Netdraw
In the window that appears search for the name of the file to open (Figure 43).
Make sure the following options are selected:
File format: → Ucinet and
Type of Data: 2 mode network → OK.
Figure 43: Find the Ucinet file
The 2-mode network (Figure 44) chart appears showing how the actors are grouped around different information demands.
Figure 44: The 2-mode network
G Gi
Figure 45: Add the attributes The network map can be enhanced by adding
attribute data. This is done in the same way as before. The .txt files are Ucinet compatible.
Go to File → Open → Vna text file → Attributes (Figure 45).
Ucinet can also be used to create an attributes database. Both sets of nodes can be given attributes but only one attribute database can be used at a time.
A window will open from which to select the attribute file created in the same way as for a 1 mode network (Figure 46).
Make sure to choose:
File form: Vna and Type of data:
Node Attribute(s) → OK.
Figure 46: Find the correct file
Once this information has been added the node colour can be changed in exactly the same way as shown for 1 mode networks.
Go to Properties → nodes → color → attribute based (Figure 47).
Figure 47: Change the colour of the nodes
In the window that appears to the right of the screen select the attribute which will be used to colour the nodes. In the example in Figure 48 the different colours represent the number of cows owned by the producers interviewed.
Figure 48: Choose the colour of the nodes
Groups can be hidden depending upon their attributes to facilitate a more detailed analysis (Figure 49). As for the 1 mode network go to Nodes and unselect the group(s) of actors which should be hidden.
Figure 49: Hide groups of nodes
The programme also allows the colour of the lines to be changed to highlight the different communication media that have been suggested to receive each type of information.
Go to Properties → Lines → Color → Tie strength (Figure 50).
Figure 50: Change the colour of the ties
This function makes it possible to highlight very specific information. For example, if we only want to see what types of information people want to receive from the radio, go again to the window on the right: relationships and chose the number assigned to the radio (which for the purposes of this exercise is the number 6) and only these lines will appear.
Figure 51 shows that some of the information nodes are now isolated from the rest of the network. This means that no one has requested to receive those types of information from the radio. Different combinations of relationships can be selected depending upon the finality of the study in question.
Figure 51: Select tie variables
This new network can be given shape using the same layout tool as for a 1 mode network.
Figure 52 shows the new network structure when looking at requests for information through the radio.
Figure 52: Network structure showing only one type of tie
6. A flexible tool
As stated at the beginning of this text, the aim of this manual is to make social network analysis (SNA) widely available to people working in rural development and other sectors in which there are highly complex social networks, as these can either help or hinder the work of researchers and development projects, depending upon the relations they establish with these existing networks.
This manual can be adapted to the requirements of people and organisations working in many different circumstances. The potential variety of users in different sectors will be fundamental to updating the manual as this is a new tool whose applications still need to be explored and tested. Therefore each user should use this methodology according to their own needs and interests, identifying the how best way to use it to understand the complex social relationships that have an influence on local development. All contributions are welcome to continue streamlining and validating this methodology.
It is hoped that by simplifying SNA more users from different sectors will be able to access it as it is a flexible, practical tool that has a huge potential in the development and research sectors. It is still too early to come to any final conclusions and for the time being it is important to continue to apply this tool in a variety of different situations to further understand and evaluate both the potential and limitations of this methodology.
There is no single answer to explain the ‘why’ of poverty, nor the ‘how’ of development; understanding the combination of factors that lead to the success or failure of development projects requires flexible thinking and the ability to make constant adjustments. This manual seeks to help practitioners and researchers understand the complementarities and conflicts that exist among different actors and initiatives in the development sector in the hope that this tool will eventually take is place among the wide range of methodologies that are used to understand and encourage development processes.
Naturally those using this manual will have a lot of questions about SNA and the results it provides. This manual deliberately does not discuss how to carry out the analysis of results, so that each user is free to draw their own conclusions, reading the maps according to their own points of interest and highlighting the importance of seeing the same reality from different points of view. The concept that supports this idea is that visualising social networks can help people to understand how complex their environment is and use this as a first step towards making positive changes.
It is hoped that this manual will be a starting point which will enable users to understand the practical details of social network analysis, so that they will be free to use their imagination to explore how to use and adapt this tool to their own needs.
Good luck using this tool!