Devices & Hardware

The Language Competence Survey of Jamaica

Published
of 37
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Description
The Language Competence Survey of Jamaica DATA ANALYSIS THE JAMAICAN LANGUAGE UNIT DEPARTMENT OF LANGUAGE, LINGUISTICS & PHILOSOPHY FACULTY OF HUMANITIES & EDUCATION UNIVERSITY OF THE WEST INDIES, MONA
Transcript
The Language Competence Survey of Jamaica DATA ANALYSIS THE JAMAICAN LANGUAGE UNIT DEPARTMENT OF LANGUAGE, LINGUISTICS & PHILOSOPHY FACULTY OF HUMANITIES & EDUCATION UNIVERSITY OF THE WEST INDIES, MONA 1 September 2007 ACKNOWLEDGEMENTS The Jamaican Language Unit (JLU) wishes to thank the students of the L331 class who took part in the data collection process, the graduate students who supervised the field work and the office staff and the data entry personnel for their cooperation in making this research project a successful one. We would also like to especially thank Mr. Michael Yee-Shui who prepared this statistical report of the data analysis. 2 Table of Contents List of Tables 4 Executive Summary 5 Sample and Analytical Plan 7 Profile of the sample 7 Profile of the Interviewers and Interviews 9 Data Analysis and Manipulation 10 Data Presentation 12 Bilingualism 12 Independent Variables: Region, Urban/Rural, Age, Gender, 12 Occupation Controlling Variable: Gender of Interviewers 15 Controlling Variable: Language Used to Initiate Interview 18 Conclusion 21 Appendix 23 Questionnaire 23 SPSS output 25 3 Table 1: Table 2: Table 3: Table 4: Table 5: Table 6: Table 7: Table 8: Table 9: Table 10: Table 11: Table 12: Table 13: Table 14: Table 15: Table 16: Table 17: Table 18: Table 19: List of Tables Demographic Variables in the Survey Structure of the Stratified Sample Characteristics of Interviewers and Interviews Bilingualism Bilingualism by Region Bilingualism by Urban/Rural Bilingualism by Age Bilingualism by Gender Bilingualism by Occupation Re-examining Bilingualism by Region, Controlling for the Effects of the Gender of Interviewers Re-examining Bilingualism by Urban/Rural, Controlling for the Effects of the Gender of Interviewers Re-examining Bilingualism by Gender, Controlling for the Effects of the Gender of Interviewers Re-examining Bilingualism by Age, Controlling for the Effects of the Gender of Interviewers Re-examining Bilingualism by Occupation, Controlling for the Effects of the Gender of Interviewers Re-examining Bilingualism by Region, Controlling for the Effects of the Language Used to Initiate the Interviews Re-examining Bilingualism by Urban/Rural, Controlling for the Effects of the Language Used to Initiate the Interviews Re-examining Bilingualism by Gender, Controlling for the Effects of the Language Used to Initiate the Interviews Re-examining Bilingualism by Age, Controlling for the Effects of the Language Used to Initiate the Interviews Re-examining Bilingualism by Occupation, Controlling for the Effects of the Language Used to Initiate the Interviews 4 LANGUAGE COMPETENCE SURVEY OF JAMAICA 2006 Executive Summary In 2005, the Jamaican Language Unit (JLU) conducted its first Language Attitude Survey of Jamaica (LAS), an island-wide study, to assess the views of Jamaicans towards Patwa (Jamaican Creole) as a. This year s study: the Language Competence Survey of Jamaica (LCS) however concentrated on the ability of Jamaicans to code switch between both s, that is Patwa and English. In other words, the 2006 study sought to assess the level of bilingualism that is exhibited by Jamaicans and to delineate some of the characteristics that are important in understanding bilingualism. The parameters of the sampling methodology were more or less maintained, with one minor modification to one of the stratifying variables used for sampling in the previous year s study. Specifically, the sample consisted of 1000 Jamaicans, stratified along the variables of region (western and eastern), area (urban and rural), age groups (18-30 years, years and years), and gender. The survey methodology was modified to more of a (hybrid) quasi-experimental design rather than the standard correlational design (typical of surveys) used last year. This change in the survey design and focus necessitated changes in the approach to data analysis. Firstly, fewer relationships were examined. This was due to the 2006 survey s more specific focus, as well as the approach to measurement of bilingualism that was taken. The present study utilised three variables essentially measuring the same construct, which were combined in the data analysis to get the best measurement of bilingualism, the dependent variable. This is unlike what occurred in 2005, when several dependent variables were used as the basis for analysis. Secondly, with the design change it was considered prudent to examine potential confounding relationships. For instance there could have been an interaction between the gender of the interviewers and the willingness of respondents to exhibit bilingualism (this is only true if interview teams were randomly assigned to interviews). The results indicate that 46.4% of respondents were able to switch between both s (with and without prompting) and therefore demonstrated bilingualism. The majority of the sample however was monolingual, with more than a third of this proportion being Patwa speakers (Jamaican Language users). 5 When bilingualism was examined using the demographic characteristics of respondents there were only two significant relationships. Demonstrated bilingualism tended to be slightly higher among respondents who were from eastern parishes that were urban areas when compared to their western and rural counterparts. Among monolingual respondents, eastern and urban areas tended to have more monolingual English speakers than western and rural areas. There was also a tendency for higher skilled or professional respondents to demonstrate bilingualism than respondents who indicated that they were unskilled or unemployed. Additionally, English speaking monolinguals tended to be concentrated in the highly skilled and professional groups. There was some amount of interaction between the gender combination of the pair of interviewers as well as the in which the interviewers initiated the interview process, and the respondent s behaviour. Respondents from urban areas who had two female interviewers were more likely to demonstrate bilingualism than those from rural Jamaica, while those from the eastern region were more likely to be monolingual English speakers than those in western parishes. Additionally the relationship between Occupation and Bilingualism was significant across all levels of the control variables but the relationships were stronger for mixed gender interview teams (teams consisting of male and female interviewers) and interviews initiated in Patwa. 6 Sample and Analytical Plan In this section of the report, the demographic structure of the sample will be presented, along with how these characteristics were used to stratify the sample. The breakdown of the characteristics of the interviewers and interviews is also presented. Additionally a brief description of the analytical plan is provided, including the data manipulations, statistics used, level of significance used for testing and a simple diagrammatic presentation of the analytic procedure. Profile of the Sample Table 1: Demographic Variables in the Survey (N= 1000) Variables Frequency % Western Region Eastern Urban/Rural Urban Rural Male Gender Female Age Occupational Groups yrs yrs yrs Unskilled/Housewives Unemployed Farmers/skilled craftsmen Clerical sales/services Self employed/ service professionals As shown in Table 1, the majority of the respondents were from eastern parishes (60%) and the other 40% were pulled from western parishes. This is unlike the previous year in which the respondents were divided equally between western and central parishes. There were equal proportions of respondents from urban and rural areas compared to 3.8% more respondents from urban areas in Table 2: Structure of the Stratified Sample Region Urban/Rural Sex Age groups yrs yrs yrs Urban Males 33 (49.3%) 34 (50.7%) 32 (48.5%) Females 34 (50.7%) 33 (49.3%) 34 (51.5%) Western All Sex Rural Males 32 (48.5%) 31 (48.4%) 36 (52.2%) Females 34 (51.5%) 33 (51.6%) 33 (47.8%) All Sex All Areas Urban Males 65 (56.5%) 50 (40%) 35 (58.3%) Eastern Females 50 (43.5%) 75 (60%) 25 (41.7%) All Sex Rural Males 50 (50%) 49 (48%) 48 (49%) Females 50 (50%) 53 (52%) 50 (51%) All Sex All Areas The gender distribution has remained comparable across the two years with roughly equal proportions of male and female respondents. Last year there were slightly more men than women, this year that has been reversed, with one respondent not specifying gender. There was greater heterogeneity in the distribution of the age groups in the present sample. Last year the sample was divided roughly into thirds across the three groups. This year almost thirty five percent were between the ages of years (34.9%) and less than a third (29.3%) was in the oldest age category. The largest occupational groups were unskilled/housewives (24.6%) and farmers/skilled craftsmen (24.1%) compared to clerical sales/services (25.4%) and farmers/skilled craftsmen (23.8%) in The unemployed category (19.8%) this year is slightly larger than the 12.2% of the sample last year. The self employed/ service professionals were 16.7% of all respondents, down from 20.4% in Region (western and eastern), Urban/Rural (urban and rural), age (18-30 years, years and years) and gender were the variables used to design the stratified sample for the LCS. The resulting design had 24 distinct strata, as displayed in Table 2. For the western parishes, there were roughly equal proportions of male and female respondents across all age groups. There was greater variability in the gender and age distributions for rural as opposed to urban areas. There were greater disparities in the age and gender distribution in urban areas of the eastern parishes, actually exhibiting the greatest heterogeneity for any set of strata. The most salient feature is a 15.1% drop in the total number of respondents in the oldest age groups while the other two age groups had 5.8% and 9.2% increases in the numbers of respondents respectively, compared to the previous year. The rural parishes have a similar pattern to those of the strata for western parishes as well as the previous year and therefore there is relative uniformity in the distribution of age and gender. Profile of the Interviewers and Interviews Table 3 highlights that approximately of a third (33.7%) of the interviews were conducted by mixed gender interview couples. This was more a function of the disparities observed in the general university population (University of the West Indies, Mona campus), from which the interviewers were selected, rather than a specific design feature. There seemed to be a preference, irrespective of the gender combinations of the interviewing teams, in the used to start the interviews, the majority 9 (53.2%) of which was started in Patwa. This roughly translates into six percent more interviews initiated using Patwa. Table 3: Characteristics of Interviewers and Interviews Variables Frequency % Male & Female Sex of interviewers Female & Female Language used to initiate interview English Patwa Data Analysis and Manipulation The data was analyzed using the Statistical Package of the Social Sciences (SPSS). The variables used in the analysis were categorical, therefore the Chi-square statistic was used to examine the bivariate relationships. Additionally, all relationships were tested using a significance level of five percent (5%). The implication of this is that the maximum probability of the risk of making a Type I error was Therefore all displayed significance levels that were below 0.05 were deemed to be statistically significant (any significance level that was exactly, as well as when rounded, equal to or greater than 0.05, was considered to be statistically insignificant). Diagram 1 is the graphical representation of the analytical plan that was used in the study. On the left hand side of the diagram are the independent variables (region, area, age groups, gender and occupational groups. On the right hand side is the dependent variable (bilingualism) and the variables located at the bottom centre (gender of interviewers and interviews initiated) are the control variables. The control variables are considered to be mediating the relationships between each of the independent Diagram 1: Analytical Plan 10 variables and the dependent variable. These relationships were assessed to identify potential confounding relationships. Generally, only the relationships that were statistically significant were reported and discussed. There are two notable variable modifications that were made for the analysis. The variable used to measure occupation groups was created by recoding the variable OCCUPAT. The original variable had a total of nine categories was simply regrouped into five (which can be seen in Table 1 above). Specifically, the categories labeled self employed/service professionals, farmers/skilled craftsmen and unemployed were created by collapsing as the names suggest self employed professionals with service professional, farmers with skilled crafts men and unemployed consisted of students, retired and unemployed respondents. This was done primarily to achieve parity with what was done in the previous year as well as to subsume categories into larger operational categories for occupational groups. The variable BILINGUALISM was a proxy variable used to measure competence, was created by the summation of three variables; Q8 (Language at scenario Jamaican or English), Q9 (Language at prompt Jamaican or English) and Q10 (Language at debrief Jamaican or English). These variables were first recoded, weighting the values of each variable to ensure that each characteristic represented by these variables would be clearly distinguishable when summed. After the creation of the proxy variable it was recoded into the three groups displayed in Table 4 below. This seemingly elaborate undertaking was done because each variable (Q8, Q9 and Q10) measured different aspects of the process used to measure bilingualism. Therefore no one variable was suitable as an adequate measure of bilingualism. This then necessitated the combination of all three to develop an accurate (as was possible) measure of bilingualism. 11 Data Presentation Bilingualism Table 4: Bilingualism Variable Frequency % Monolingualism English Patwa Bilingualism Demonstrated Bilingualism From Table 4, it can be seen that 46.4% of the respondents demonstrated bilingualism. Less than 20% of the sample were monolinguals that spoke only English and just over a third (36.5%) of the respondents were Patwa speaking mono-linguals (either because they did not speak both s during the interview or told the interviewers that they were capable of doing so but did not demonstrate competence in both). Independent Variables: Region, Urban/Rural, Age, Gender, Occupation Table 5-9 present the results of the chi-square analysis, examining the relationships between bilingualism and region, Urban/Rural, age, gender and occupation. Only three of relationships were found to be statistically significant, namely Region, Urban/Rural and Occupational Groups with Bilingualism. Table 5: Bilingualism by Region Variables Bilingualism Region English Patwa Demonstrated Bilingualism χ 2 = 7.998, p = (%) (%) (%) Western 54 (13.5%) 162 (40.5%) 184 (46%) n = 400 Eastern 117 (19.5%) 203 (33.8%) 280 (46.7%) n = Table 6: Bilingualism by Urban/Rural Variables Bilingualism Area English Patwa Demonstrated Bilingualism χ 2 =11.365, p =0.003 (%) (%) (%) Urban 103 (20.6%) 163 (32.6%) 234 (46.8%) n = 500 Rural 68 (13.6%) 202 (40.4%) 230 (46%) n = 500 Table 7: Bilingualism by Age Variables Bilingualism Age Groups English Patwa Demonstrated Bilingualism χ 2 =4.978, p = (%) (%) (%) yrs 69 (19.8%) 115 (33%) 165 (47.3%) n = yrs 60 (15.7%) 142 (37.1%) 181 (47.3%) n = yrs 42 (15.7%) 108 (40.3%) 118 (44%) n = 268 Table 8: Bilingualism by Gender Variables Bilingualism Gender English Patwa Demonstrated Bilingualism χ 2 = 0.074, p = (%) (%) (%) Male 86 (17.4%) 181 (36.6%) 228 (46.1%) n = 495 Female 85 (16.9%) 183 (36.3%) 236 (46.8%) n = 504 Table 9: Bilingualism by Occupation Variables Bilingualism Occupational Groups English Patwa Demonstrated Bilingualism χ 2 = , p = (%) (%) (%) Unskilled/housewife 21 (8.5%) 127 (51.6%) 98 (39.8%) n = 246 Unemployed 45 (22.7%) 66 (33.3%) 87 (43.9%) n = 198 Farmer/skilled 28 (11.6%) 100 (41.5%) 113 (46.9%) n = 241 craftsman Clerical 25 (16.9%) 36 (24.3%) 87 (58.8%) n = 148 sales/services selfemployed/service professional 52 (31.1%) 36 (21.6%) 79 (47.3%) n = Region There was a statistically significant relationship between Region and Bilingualism (χ 2 (4) = 7.998, p 0.05). As shown in Table 5, there was a marginal difference in the number of bilinguals across the regions: eastern parishes had 46.7% compared to 46% in the western parishes. Among monolinguals, it would appear that respondents who were from eastern parishes (19.5%) were more likely to exhibit English monolingualism than those from western parishes (13.5%). The reverse is true for monolingual Patwa speakers, where 40.5% were to be found in western parishes compared to a third in eastern parishes. There was a very weak association between the two variables (cc = 0.089), with less than one percent of the variation in bilingualism being explained by its relationship with region. Urban/Rural The results indicate that a statistically significant relationship exists between Urban/Rural and Bilingualism (χ 2 (2) = , p 0.05). Respondents from urban areas were less likely to be Patwaspeaking mono-linguists (20.6%) and fractionally more likely to demonstrate bilingualism (46.8%) when compared with persons from rural areas (13.6% and 46%) respectively. There was a weak relationship between area of residence and bilingualism (cc = 0.106). Additionally, approximately one percent of the variation in the distribution of Bilingualism was explained by its relationship with area. Occupational Groups In terms of the relationship between Occupation and Bilingualism, there was direct variation between occupational classification groups and being an English speaking monolingual or exhibiting bilingualism. That is, as the level of skill (or education required) for the job increased or the occupational categories become more service oriented, respondents were more likely to either be English-speaking monolingual or be bilingual rather than a Patwa-speaking monolingual. From Table 9, it can be seen that unskilled workers or housewives (51.6%) were most likely to demonstrate Patwa monolingualism. Clerical sale/ services and self employed/ service professionals were most likely to demonstrate bilingualism (58.8% and 47.3% respectively). There was a weak relationship between the two variables (cc= 0.271), with 7.3% of the variation in bilingualism being explained by its relationship with occupational groups. 14 Controlling Variable: Gender of Interviewers Tables 10 to 14 present the results of the chi-square analysis examining the relationships between bilingualism and the independent variables (Region, Urban/Rural, Age, Gender and Occupation), controlling for the effects of the gender of the interviewers. As before, the only significant relationships were found between Region, Urban/Rural and Occupational Groups. Table 10: Re-examining Bilingualism by Region, Controlling for the Effects of the Gender of Interviewers Gender of Bilingualism Western Eastern Interviewers count(%) count(%) Male & Female English 11 (8.5%) 43 (20.7%) χ 2 = 8.905, p = Patwa 50 (38.8%) 74 (35.6%) Demonstrated Bilingualism 68 (57.7%) 91 (43.8%) Female & Female χ 2 = 4.967, p = English 43 (15.9%) 74 (18.9%) Patwa 112 (41.3%) 129 (32.9%) Demonstrated Bilingualism 116 (42.8%) 189 (48.2%) Table 11: Re-examining Bilingualism by Urban/Rural, Controlling for the Effects of the Gender of Interviewers Gender of Bilingualism Urban Rural Interviewers (%) count(%) Male & Female English 31 (20.5%) 23 (12.4%) χ 2 = 4.468, p = Patwa 55 (36.4%) 69 (37.1%) Dem
Search
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks