Jun. 17, 2009 - Automatic coding |
| In MAXQDA are the references to a query automatically one or more arbitrary category. This process is automatic encoding and is especially good for a quick Explo-ration suitable, but three points are observed. First, the context dependence to be considered a search, because it makes a difference if someone says I have the textbook used or ichhabe the textbook is not used. Secondly, can anyone on the textbook content to speak, even if the word is not used. And thirdly, a misspelled word is simply automatic sein.Auch if no single code of computer-based analysis of qualitative data, it is extremely useful, not only of heuristic value (see Kuckartz 2007: 126). We have the auto-code function, for example, for the categorization of the question used the role of Statistics Textbook by Jürgen Bortz for the preparation of the substance has played. With the additional program MAXDictio 5, we could begin a list of all words in the students create the final answer to all the spellings of the word "Bortz" from resourceful to make. For example, the students also had "Borzt" or "Borz" in their written answers. After the survey research design automatic code Zuordnungdes "substance preparation> Bortz" to the respective paragraphs, we have a total of 66 responses from 64 students and could durchgeschaut-ing seven categories inductively generate the handling and use of the textbook to reflect. The results of the automatic coding must normally by manual control at überprüftwerden sense: As the students Bortz not only in relation to the substance called mastery, we have more irrelevant Codie requirements deleted. |
| • Comments • Post A Comment! • Permanent Link |
Jun. 17, 2009 - Evaluate Data Processing |
For the presentation of the fifth step, we draw out three aspects to the computational approach particularly interesting: the Codierregeln the automatic encoding data processing and the transformation of variables in the code. CodierregelnRegeln, after which the Codiervorgang from equip to do this, perform various functions. From a pragmatic viewpoint simplify A subsequent analysis, because it would be very cumbersome, even if only one word and another, a whole sentence would be encoded. From a methodological perspective, serve Co-dierregeln in particular for quality assurance. For our evaluation study we had to lay down rules, inter alia, the very different behavior of the response that students consider:
1. Scope of Encodings: There will always be the entire answer to a ques-coded ge. If the answer is more than three lines or more than one, thematically different sentences, only the encoded part of the category is relevant. Note: Unlike personal interviews can be found on the answers entered on the computer rarely finely rates, but dieAntworten vary from individual words on bulleted lists, to multiple sets. Accordingly, differ also in terms of their length encodings.
2. Double encodings: it is an information only once per case coded. If someone, for example, on several occasions to noise in the lecture indicates, this is a fact only once encoded. Note: This rule, we have positioned ourselves as important was to find out how many people such as noise and do not designate how often he referred to as a whole is.
3. Distributed encoding: There are always answers all of the people be taken into account. If a person in the question of what it has fallen ill, including a suggestion for improvement indicates, this is also in the corresponding coded question - unless he already has been encoded and it is a double-coding would act. Note: This rule ensures that all information about a topic also be encoded. The easiest way is all the "extraneous information" zueiner the same question in the correct category mitzucodieren, because you do not always understand all the answers should be. It is important to mehrerenCodierern that advance a procedure is discussed, as in a team with such a distributed encodings deal.
4. Treatment of missing values ( "Missing") Blank answers or responses, which are empty to indicate ("-"), in the subcategory "Mis-sing" in the respective issues. The answer "no" will be out against a sub-category, for example, when asked what the students were bad.
Note: This rule is used in subsequent enumerations of Codehäu-figkeiten to identify the persons that are not on a question to have answered-tion. In this respect, the online survey of open issues from personal interviews, in which Missingsummary Hardly to be expected. |
| • Comments • Post A Comment! • Permanent Link |
|
About Me
Survey research design advice
Friends
|
