A standard procedure in conducting a consistent laddering interview =================================================================== Consistent Laddering seeks to do the same as Hinkle's (1965) laddering technique, just a bit more extensively. The main point of difference is that elements must be rated on all elicited constructs, and ties that connect constructs are weighted by subjective importance in the context of their direct successors in the network of constructs. Other than that, the procedure is very similar to traditional laddering procedure. The standard procedure using CLAD consists of the following: Create a new project and interview --------------------------------- Agree on the topic of the interview with the participant. The boundaries of the topic should be limited to one or more guiding phrases. Otherwise, the scope of the interview may be too broad. Elements considered in this interview should also be agreed with the participant. After you've created a new project in CLAD, and created a new interview within this project you are ready to start the actual interview. Elicit the initial constructs ----------------------------- After the participant has agreed to the elements, you would probably want to begin with eliciting initial constructs from elements. I recommend using file cards, wherever possible to help the participant with the elicitation procedure; this is just a suggestion. Elicit the laddered constructs ------------------------------ You can start laddering up procedure by asking the "Why?" type of questions. When a more superordinate construct is elicited, you can add it to CLAD using the "Add construct" command at the right side of the page. In a similar way a more subordinate construct, elicited by asking “In which way” type of question, can be added to the network. Rate elements on elicited constructs ------------------------------------ After a new construct has been elicited, you would probably want to rate elements on it. Normally, you'd have to select the elicited construct first, using the “Select a construct” command, but CLAD automatically selects the construct that has been added to the list of constructs most recently. In most cases, you can, therefore, skip the step of selecting the construct, and go straight to rating elements on (auto)selected construct. .. image:: _static/Add-ratings.png :align: center Add ties between constructs --------------------------- One of main concerns of laddering is the position of a construct within a hierarchically organized network of constructs. You have to add ties between constructs to indicate that. To add a tie you have to select source and destination from two lists of constructs elicited in the interview (see the picture below). In contrast to the traditional approach to laddering you can weight ties between constructs; you can ask the respondent how important a construct is in the context of a direct successor to find this out. .. image:: _static/Add-tie.png :align: center Visualize the network --------------------- Since it is very easy to get lost in a network of constructs, a standard procedure would include an occasional trip to visualization of the network of constructs. There are quite a few ways in which CLAD can visualize the network, or a part of it; weakly connected components of the graph, paths between two constructs, etc. .. image:: _static/Weak-component.png :align: center Please note that most often the visualization is done with consideration to the selected construct, e.g. CLAD visualizes the weakly connected component of the graph that contains the selected or (auto)selected construct. .. image:: _static/Path-two-constructs.png :align: center Check the data for consistency ------------------------------ The standard procedure consists of occasional checks of the consistency of network of constructs. These checks are to be used to adapt the flow of the interview and focus on subsets of constructs to see if the participant was properly understood, or the data were entered correctly, etc. .. image:: _static/EA.png :align: center Also keep in mind that what you are getting in an interview are only verbal labels of constructs that the participant uses in construing the considered topic. Inconsistencies found in the interview may also indicate that verbal labels were not properly defined, or that the participant interpreted these labels in various ways in the interview, etc.