节点文献

大学生健康生活方式评价量表研制及初步应用研究

Development and Preliminary Application of a Healthy Lifestyle Questionnaire(HLQ) for College Students

【作者】 王冬

【导师】 陈清;

【作者基本信息】 南方医科大学 , 流行病与卫生统计学, 2009, 博士

【摘要】 背景:生活方式影响个人健康已成为全球瞩目的焦点,WHO指出:“每个人的健康与寿命60%取决于行为与生活方式”。良好的健康生活方式有赖于早年生活习惯的养成,成年人很难去改变年轻时养成的不健康生活习惯,要提高人们的健康生活方式水平,重点应放在年轻人。大学阶段是个人逐渐独立和健康行为发展定型的关键阶段,又是接受知识和塑造行为的黄金时期,此时期的生活方式会延续到成年,影响一生的健康。目前国内大陆对大学生生活方式的研究重理性思辨,轻实证研究,多集中于健康危险行为,尚停留在行为与健康的关系上,针对大学生健康生活方式测量的研究工具未见报道,也很少对影响大学生健康生活方式的因素进行实证研究。因此基于量化我国大学生健康生活方式,探索影响其健康生活方式养成的因素,而开展本课题研究。目的:1.充分考虑我国大陆文化背景、社会结构和价值观,借鉴国外和台湾地区生活方式评价工具,采用系统量表开发技术,编制大学生健康生活方式评价量表,为标准化评估大学生健康生活方式提供客观有效的工具。2.通过评价量表的应用,了解大学生健康生活方式的现状,初步探索影响大学生健康生活方式的因素,从学校角度提出促使大学生养成健康生活方式的建议对策,为高校倡导大学生养成健康生活方式提供参考。方法:1.采用文献研究法、头脑风暴法、焦点小组(大学生)评价法、专家咨询与访谈法编制大学生健康生活方式评价量表条目池。2.根据专业声誉及研究方向,选取相关领域39名专家,采用Delphi法对备选条目池进行3轮初筛,形成大学生健康生活方式评价量表的初稿。3.按照地理位置,在东西南北中五个区域各取一所高校,每所高校取样100人进行预调查,采用变异系数法、评价指标的应答率、因子分析法、主成分分析法、评价条目间以及评价条目与其维度的相关性分析等方法筛选条目,形成大学生健康生活方式评价量表测试版。4.按照高校类别抽取4所院校在校大学生2000人进行现场调查,从信度、效度、反应度等方面对大学生健康生活方式评价量表测试版进行考评。5.采用专家咨询法,进一步修订测试版,完成大学生健康生活方式评价量表的编制。6.运用自编大学生健康生活方式调查问卷对两阶段分层抽样抽取的9所院校大学生进行现场调查,采用单因素与多因素分析初步探索影响大学生健康生活方式的因素。结果:1.大学生健康生活方式评价量表条目池综合专家和学生对大学生健康生活方式条目重要性评价结果,按预想结构归类成由外显性和内隐性健康生活方式2个方面,13个维度,117个条目构成的大学生健康生活方式评价量表条目池。2.大学生健康生活方式评价量表初稿(1)第一轮征询结果:专家的积极系数为84.6%;33位专家对各条目评价的权威系数的均数在0.577~0.818之间,平均为0.737;专家对各条目重要性判断的总协调系数为0.212,各维度协调系数在0.083~0.360;按照重要性转换值均数大小排秩,结合专家对条目修改、合并及删除意见,得出8个维度,70个条目。(2)第二轮征询结果:专家的积极系数为93.94%;31位专家对各条目评价的权威系数的均数在0.73d~0.945之间,平均为0.834;外显性和内隐性健康生活方式相对重要性评价为0.580:0.420;专家对各条目重要性判断的总协调系数为0.317,各维度协调系数在0.217~0.402;得出外显性和内隐性健康生活方式2个方面,8个维度,44个条目。(3)第三轮征询结果:专家的积极系数为96.77%;30位专家对各条目评价的权威系数的均数在0.778~0.903之间,平均为0.850。外显性和内隐性健康生活方式相对重要性评价为0.583:0.417;专家对各条目重要性判断的总协调系数为0.401,各维度协调系数在0.298~0.526;得出外显性和内隐性健康生活方式2个方面,7个维度,42个条目。3.大学生健康生活方式评价量表测试版问卷回收有效率94.4%,经过六种方法对42个初选条目进行了再筛选,筛选出由外显与内隐性健康生活方式2个方面,7个维度,40个条目构成的大学生健康生活方式评价量表测试版。4.大学生健康生活方式评价量表测试版考评(1)信度考评结果大学生健康生活方式评价量表7个维度及外显性、内隐性和总的健康生活方式首次测试得分与间隔二周测试得分相关系数在0.501~0.842之间(p<0.001);总的分半信度相关系数0.865(p=0.000),各维度分半信度在0.422~0.746之间;总的Cronbach’α系数为0.898,除健康危害行为维度的Cronbach’α系数为0.417外,其它维度的Cronbach’α系数在0.607至0.910之间。(2)效度考评结果量表构成能够反映大学生健康生活方式的真实情况,支持内容效度;大学生健康生活方式评价量表测试版总分与WHO生活质量测定量表简表得分总分的相关系数0.571:采用因子分析评价结构效度,KMO统计量为0.932,Bartlett’s球形检验结果χ~2=20883.975(p=0.000),适合于因子分析,经因子分析,得到9个公因子,其累计贡献率为54.37%。各因素与量表的设计构想不同,结构效度较不理想,表明本量表需要进一步对条目结构进行修订。在40个集合效度试验中,成功率为97.5%,在240个区分效度定标试验中,成功率为100%,同时各维度与其它维度间的Spearman相关系数均小于该维度的Cronbach’α系数;各维度得分与总得分间的相关系数均大于各维度得分间的相关系数。各维度分与其分量表分的相关系数均较大,而与其它分量表分相关系数较小,这些均强有力地支持量表的集合和区分效度。(3)反应度考评结果大学生健康生活方式评价量表测试版各维度首次测试得分与间隔二周测试得分相关系数在0.582~0.843之间(p<0.05);经配对t检验,两次测试结果间除饮食营养行为维度外,其余均无显著性差异(p>0.05);大学生健康生活方式评价量表测试版各维度首次测试得分与间隔四周测试得分相关系数在0.526~0.853之间(p<0.05);经配对t检验,两次测试结果间除运动锻炼行为维度外,其余均无显著性差异(p>0.05),显示量表的反应度较好。(4)人口统计因素考评结果:大学生健康生活方式评价量表测试版的总分在院校类别、家庭月收入、父母亲文化程度、入学前居住地和是否独生子女等人口统计因素方面存在差异,在性别和年级方面不存在差异。5.大学生健康生活方式评价量表大学生健康生活方式评价量表由8个维度,38个条目组成,涉及到外显和内隐两个方面生活方式,其中1至22条目组成大学生外显性健康生活方式评定分量表,23至38条目组成大学生内隐性健康生活方式评定分量表。5.大学生健康生活方式评价量表初步应用(1)大学生健康生活方式评价量表的考核①信度按Spearman-Brown校正,得全量表的分半相关系数为0.840,二个方面的分半相关系数分别为0.710,0.808,8个维度的分半相关系数在0.487~0.737之间;全量表的Cronbach’α系数为0.892,二个方面系数α分别为0.770和0.891,8个维度的α系数在0.518~0.830之间。②效度KMO统计量为0.929,Bartlett’s球形检验结果χ~2=54913.124(p=0.000),适合于因子分析。经因子分析,得到9个公因子,累计贡献率为55.02%。除健康危害行为分为两个因子外,其余各条目在因素上的负荷与量表结构吻合,支持结构效度。在38个集合效度试验中,成功率为97.37%;在259个区分效度定标试验中,成功率为97.30%;各维度与其它维度间的Spearman相关系数均小于该维度的Cronbach’a系数;各维度得分与总得分间的相关系数均大于各维度得分间的相关系数。各维度分与其分量表分的相关系数均较大,而与其它分量表分相关系数较小,这些均支持集合和区分效度。③大学生健康生活方式评价量表的反应度大学生健康生活方式得分能够区分出健康大学生与患慢性病大学生,经常、一般和很少生病/身体不舒服的大学生。(2)大学生健康生活方式及影响因素初步分析①大学生健康生活方式总得分标准分为67,接近中等程度,在各维度上,大学生健康责任行为得分最高,生命欣赏行为次之,其余依序是人际支持行为、生活规律行为、压力管理行为、健康危害行为、饮食营养行为,而运动锻炼行为最低。②性别、气质、自感健康、自感成绩、院校所在地经济水平、院校类别、年级、和谁一起居住、是否学过健康教育与健康促进课程、出生地经济水平、家庭月收入和父母亲文化程度等影响大学生健康生活方式。结论:1.本研究研制的大学生健康生活方式评价量表包含运动锻炼行为、生活规律行为、饮食营养行为、健康危害行为、健康责任行为、人际支持行为、压力管理行为、生命欣赏行为等8个维度,38个条目。2.本量表有较高的信度和效度,可以作为测量大学生健康生活方式的工具。3.大学生健康生活方式接近中等程度,大学生虽然有较高的健康责任,但较少从事运动锻炼行为。4.影响大学生健康生活方式的因素多且复杂,包括生理、心理及认知、学校因素和家庭等多方面的因素。学校应根据大学生的这些特点采取相应的对策激发不同学生采取健康生活方式的兴趣,促使大学生养成健康生活方式。

【Abstract】 Background:During recent decades, lifestyle is an important determinant of health status, which has become a focus that attracted increasing interests around the world. The World Health Organization pointed out that 60 percent of the quality of an individual’s health and life depends on his behavior and lifestyles. A good health-promoting behavior depends on the living habits adopted in the early years of one’s life. It is far more difficult for adults to change unhealthy habits adopted in their youth, so the young people are the major population whose healthy lifestyles should be attached great importance to.College student life is an important stage when individual independence and healthy behavior are to be established, and also a golden period when knowledge is absorbed and personality is shaped. Meanwhile, during this stage, unhealthy practices and behaviors will also continue into adulthood and thus jeopardize their health status in later life. At present, In China Mainland, the research on the life styles of college students is mostly focused on the health risk behavior and remains at research phase on the relationship between behavior and health, namely, the research only addresses the life styles as factors that affect health. So far the relevant research approach aiming at the measurement of the college students’ healthy lifestyles have not been reported; few studies on the factors that affect healthy lifestyles of college students have been carried out, and almost no empirical study thereof has been considered. Therefore, based on this, the present research attempts to focus mainly on the quantification regarding college students’ healthy behaviors so as to explore the influential factors on their healthy life style-forming.Objectives:1. To provide the objective, effective and standard tools to evaluate the healthy lifestyles of college students, this research fully considers, by adopting the tools on evaluating life styles in Taiwan region and abroad, the cultural background, social structure and values of young people in China Mainland. Besides, the research used the systematic technology of developing scale to generate the healthy lifestyle questionnaire (HLQ) for the college students.2. Through HLQ, the current situation of the college students’ lifestyles will be understood, and the relevant influence factors from many aspects will be explored. Apart from this, the strategies to promote healthy lifestyles of college students will be proposed to serve as the reference for the departments of educational administration to promote the college students’ lifestyles.Methods:1. An item pool of HLQ for College Students is to be complied based on literature reviews, brain storm, focus subject (undergraduates) interview, and expert consultation and interview.2. Totally 33 experts based on their professional reputation and research domain were selected to form a panel and the 3 round Delphi method was adopted to form the first draft of HLQ.3. Totally 500 college students from 5 universities located separately in the eastern, western, southern, northern, and middle China were chosen as subjects to investigate via questionnaire, disperse trend, factor analysis, response rate, principal component analysis, correlation coefficient among and between item and domain were adopted to further screen the items for HLQ.4. The validity, the reliability and the responsiveness of the test version of HLQ were worked out from the data of totally 2000 college students selected from four universities.5. The method of expert consultation was adopted to update the test version of HLQ.6. Totally 5400 students from nine universities were selected by a two-stage stratified cluster sampling method. These subjects were also surveyed with the HLQ and self-designed questionnaire containing demographic variables and the other factors by the method of univariate analysis and multivariate analysis.Results:1. The item pool of HLQ for College studentsThe item pool of HLQ consists of 117 items and was classified into 2 sub-scales concerning the extrinsic healthy lifestyles and implicit healthy lifestyles and 13 domains.2. The first draft HLQ for College students2.1 Result of the first round consultationThe positive coefficient of experts is 84.60%. The means concerning authority coefficient of each item evaluated by 33 experts range from 0.517 to 0.818, and the average is 0.737. The total coordination coefficient (Kendall’W coefficient) of each item evaluated by 33 experts is 0.212, and the coordination coefficient of each domain goes between 0.083 and 0.360, The 70 items falling into 8 domains were obtained according to the mean of transformed importance value and experts’ updated suggestion on the items.2.2 Result of the second round consultationThe positive coefficient of experts is 93.94%. The means of authority coefficient of each item evaluated by 31 experts go between 0.730 and 0.945, and the average is 0.834. The relative importance of the extrinsic healthy lifestyles and implicit healthy lifestyles is 0.58:0.42, and the Kendall’s Wcoefficient is 0.742. The total coordination coefficient of each item evaluated by 31 experts is 0.317, and the coordination coefficient of each domain is found between 0.217 and 0.402. The 44 items are observed to belong to 8 domains. Two aspects were obtained according to the mean of transformed importance value and experts’ updated suggestion on the items. 2.3 Result of the third round consultationThe positive coefficient of experts is 96.77%. The means of authority coefficient of each item evaluated by 30 experts are found between 0.778 and 0.903, and the average is 0.850. The relative importance of the extrinsic healthy lifestyles and implicit healthy lifestyles is 0.583:0.417. The total coordination coefficient of each item evaluated by 30 experts is 0.401, and the coordination coefficient of each domain is between 0.298 and 0.526. The 42 items are categorized into 7 domains. Two aspects were obtained according to the mean of transformed importance value and experts’ updated suggestion on the items.3. The test version of HLQ for College studentsTotally 472, accounting for 94.4%, valid questionnaires were used for data analyses in this study, and the test version of HLQ is composed of 7 domains and 40 items obtained from further screen by six methods.4. The HLQ for College students4.1 The appraisal result of reliabilityCronbach a of total score was 0.898 and that of 7 domains were observed to range from 0.607 to 0.910 except for health risky behavior. Retest reliability of total score is 0.708 and that of 7 domains is 0.501~0.842 (p<0.001). The correlation coefficient of Spearman-brown split-half was 0.865 and hat of 7 domains goes from 0.422 to 0.746 (p<0.001).4.2 The appraisal result of validityThe content validity was supported by prophase research, and the items of the HLQ can reflect the real present situation of healthy lifestyles of college students. The Pearson’s correlation coefficient between the total score of the HLQ test version and the total score of WHOQOL—BREF is 0.571; therefore, the criterion validity was accepted. The construct validity in factor analysis did not go well with the structure of the HLQ. The KMO statistic of 0.932 and Bartlett’s test of sphericity with x~2=20883.975 (p=0.000) suggested that the factor analysis could be performed appropriately. Both convergent validity and discriminating validity were strongly supported by the correlations within the factors which were stronger than that between the factors.4.3 The appraisal result of responsivenessThe correlation coefficient of the total score of the HLQ test version obtained between the first test and the test two weeks later is 0.582~0.843(p<0.05), suggesting higher correlation, and two test results have no significant differences except for nutritional behavior. The correlation coefficient of the total score of the HLQ test version obtained between the first test and the test four weeks later is 0.526~0.853(p<0.05), suggesting higher correlation, and the two test results indicate no significant differences except for exercise behavior. This shows good responsiveness.4.4 The appraisal result of Demographic variablesThe total score of the HLQ test version show significant differences in type characteristic of the university, family’s monthly earning, parental educational level, residence before being enrolled into university and family with one-child or not. The total score has no significant differences in terms of gender and grade.4.5 The HLQ for College studentsThe HLQ for College students is composed of 8 domains and 38 items and involvesin the extrinsic healthy lifestyles and implicit healthy lifestyles.5. The preliminary application study of HLQ for College students5.1 Appraisal of the HLQ for College students5.1.1 The appraisal result of reliabilityCorrected in accordance with Spearman-Brown, the total split-half correlationcoefficient became 0.840; two aspects split-half correlation coefficients were0.710,0.808, and that of 8 domains were 0.372~0.737. Cronbach a coefficient oftotal score was 0.892; two aspects of coefficient were 0.770 and 0.891, that of 8domains were 0.418~0.830 (p<0.001). 5.1.2 The appraisal result of validityFactor analysis used to evaluate construct validity, KMO statistic was 0.929; Bartlett’s test results of spherical x~2=54913.124 (p=0.000), suggested that the factor analysis could be performed appropriately. Through factor analysis, the nine common factors were worked out, and the cumulative contribution rate was 55.02%. In addition, health risk behavior is divided into two factors, the remaining entries in the load factors found in line with the scale structure, supporting the structural validity. At the 38 convergent validity tests, the success rate was 97.37%. At the 259 discriminating validity calibration trials, the success rate was 97.30%. The Spearman correlation coefficient between the domains was smaller than that of the Cronbach’α. The scores between the domain scores and total scores were greater than those of the domains. Comparatively speaking, the correlation coefficients of every dimension and its sub- scale were large, and other sub-scale correlation coefficient smaller. All this supports convergent validity and discriminate validity.5.1.3 The appraisal result of responsivenessThe HLQ is able to distinguish students who are healthy, students who sutler fromchronic diseases, students who are often ill and students who are physicallydiscomfort. Therefore, it demonstrates reliable responsiveness.5.2 A preliminary explorative study of influential factors on college students’ healthylifestyles5.2.1 The standard total score of college students’ healthy lifestyles is 67, near the middle level. In every domain, the maximal one is health responsibility behavior, the next one is life appreciation behavior. The rest goes in the order of significance: interpersonal supporting behavior, law of life behavior, stress management behavior, health risk behavior and nutrition behavior, the minimal one is exercise behavior.5.2.2 Some factors affect the healthy lifestyles of college students, such as gender, temperament, self-rated health, self-rated achievement, place of colleges, type of college, grade, residential peer, birth region, educational levels of parents and family’s monthly-earning. Conclusions:1. The HLQ for college students includes exercise behavior, law of life behavior, nutrition behavior, health risk behavior, health responsibility behavior, interpersonal support behavior, stress management behavior and life appreciation behavior.2. The HLQ for college students has high reliability and validity by further statistical verification to indicate that it could serve as a measuring tool for evaluating college student’ healthy lifestyles.3. The college students healthy lifestyles is close to the middle level, and have a good health responsibility behavior, but they are less engaged in exercise.4. The factors including physiological, psychological and cognitive, school and family affect the college student healthy lifestyles, and the factors are very complex, and all mutually affect the healthy lifestyles of college students. It is strongly suggested that the colleges should take different measures to stimulate student interest in taking health-promoting lifestyles based on their characteristics so as to develop healthy lifestyles during their college student life.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络