Data & data collection
- 1. Seliger and Shohami. 1989
- 2. • Data analysis or data collection consists in organizing, summarizing and synthetizing the data to get the results. • The data analysis technique will depend on the research problem, the design chosen and the type of data collected
- 3. Data collection in quantitative research In quantitative research the data will be numerical or the data will be converted in numbers and the analysis will need statistics Qualitative data will deal with non-numerical data. Most likely, in linguistic units of oral or written speech.
- 4. It is usually taught that not using statistics makes the research easier to conduct. On the contrary, using them makes the research more manageable. There are many statistical packages which can help to manage the data.
- 5. Data collection in qualitative research Qualitative requires intuition and understanding concerning the data. It is a more complex task. There are several assumptions related to using parametric statistics. They are not strong enough to reject a whole hypothesis. One assumption is that the variable studied is normally distributed in the population. Since most variables are normally distributed the assumption is usually met. A second one is that data represent an interval or radio scale of measurement. Measures used in second language acquisition represent interval data. The assumption is also usually met. The third one is that subjects are selected independently to study. So, selection of one subject does not affect the selection of others. This will be the case whenever the sample is randomly selected.
- 6. Qualitative requires intuition and understanding concerning the data. It is a more complex task. Two main techniques can be identified when analyzing qualitative data. Deriving a set of categories for dealing with text segments. This is a procedure merely inductive. Once the categories have been set, they are applied to the reminder data, which leads to the refinement of categories and the discovery of new patterns. This type of research is descriptive and exploratory. These studies are more confirmatory and aim at some kind of explanation.
- 7. Descriptive research Data is generally analyzed with descriptive statistics. These will provide information such as how often certain language phenomena occur, typical use of language elements, and the relationship between variables, among others. The types of statistics used in descriptive research are frequencies, central tendencies and variabilities. The Frequencies are used to know how often a phenomenon occurs and are based on counting the number of occurrences. Very useful in second language acquisition research, where the main interest lies in how often elements of language are used. They also provide information about the performance of the subjects on tests and questionnaires before the results are used for analyzing the data.
- 8. Correlational data It is obtained from the descriptive research and examines the relationship between variables without manipulate them. Multivariate data Obtained from the multivariate research. It can be analyzed through a set of techniques where a number of dependent and independent variables are analyzed simultaneously. These techniques can be applied when researching language aptitude, personality or learner’s background.
- 9. There are three multivariable procedures. The first is multiple regression. This permits examine the relationship and predictive power of one or more independent variables with the dependent variable. The second is discriminant analysis. This is about which combination of independent variables distinguish the most between two or more categories of the dependent variable. An example could be male/female, monolingual/bilingual, etc. The third is the factor analysis. This helps the researcher to manage larger sets of data by identifying factors that underline the data. This type of analysis is based on the assumption that variables that measure the same factor will be highly related, while the ones which measure different factors will have low correlations. This kind of analysis has been used in second language learning to validate factors that are believed to underline different language constructs such as proficiency, aptitude and attitude to learn the second language.
- 10. Experimental data In experimental research, when two groups (experimental and control) are being compared, the researcher will use something called the t-test. This is used to compare the means of two groups. The results gotten with this test are called t-value. That value is entered in t-values chart. One way analysis of variance is another technique used to collect experimental data. This analysis is performed on the variance of the groups and is focused on whether the variability between the different groups is greater than the variability within each of the groups. The F value is the radio of the between variance over the within variance. A significant F will occur when the variability among the group is greater that the variability within the group.
- 11. Chi square This data analysis procedure helps the researcher to address questions about relations between two nominal variables. During the procedure, the researcher compares the frequencies observed in a sample with the expected frequencies. Using the computer for data analysis Most of the data analysis techniques described here can be performed with the computer. There are many packages which can help you to do it. Nevertheless, it is important to know how to use them in order to have good results. A computer analysis must be planned and attention to small details must be given.
- 12. Daniel García García Mat. 10-11470