Quantitative research in language teaching and learning

开篇：润墨网以专业的文秘视角，为您筛选了一篇Quantitative research in language teaching and learning范文，如需获取更多写作素材，在线客服老师一对一协助。欢迎您的阅读与分享！

Definition of quantitative research

Quantitative research is an approach that seeks to determine the relationship between variables and cause and effect relationships. Normally， when researchers choose research questions they intend to investigate further， they identify hypotheses through literature review of other researchers’former studies. Literature review will help researchers to see if their research problems have already been done， and if there is still a gap. Thus their research hypotheses are usually built on past research to fill the gap. It is the research hypothesis that directs the researcher to the problem that needs to be solved.

Features of quantitative research

The purpose of quantitative research is to study relationships and cause and effect. The reasoning of quantitative research is deductive or top-down. The researcher uses standardized measurement and other quantitative instruments to measure what is observed. Data collected in quantitative are numbers and statistics. After collecting the data， the researcher uses statistical criteria or statistical means to analyze the data. Both the standardized instruments and statistical criteria can enhance the objectivity of the researcher. Quantitative researchers attempt to avoid the human bias and operate under the assumption of objectivity. Kervin， Vialle， Herrington， & Okely （2006） claim that “for the quantitative researcher， an objective approach is essential”

Strengths of quantitative research

There are many advantages of quantitative research. Firstly， it provides precision and control in the research that gives confidence in the result. Two reasons are contribute to this strength. It is expected that quantitative researchers set up a carefully controlled situation that can insure that no other variables will influence the result. Another is that the result of quantitative research is generalized via statistical data which make the result easy to see and more supportive. Another strength is that it is useful for studying large numbers of people. Based on large and random samples， quantitative research is relatively more useful for making generalizations about populations （Johnson & Christensen， 2004）. Moreover， quantitative research is more useful in educational context， because it can provide an insight cause for the result that allows the educationists to be much more confident to improve aspects of school performance （Kervin， Vialle， Herrington， & Okely， 2006）.

Limitations of quantitative research

However， quantitative research does have limitations. Firstly， it is not enough for the quantitative research to explore human behaviors which are more complex. Quantitative research is too mechanism to examine insight the complexity of human behavior （Lincoln & Guba， 1985）. Secondly， it is difficulty to apply the result of quantitative research directly to specific contexts and situations， because knowledge produced might be too abstract and general （Johnson & Christensen， 2004）.

1.Experimental research

Experimental research is to study cause and effect relationships between variables. In educational research， language researchers often intend to explore the relationship between a variable such as a teaching method and another variable， such as language proficiency （Nunan， 1992）. The teaching method here is called independent variable and the test scores of language proficiency is called dependent variable. The basic characteristics of experimental research are manipulation or control and random assignment. Campbell & Stanley （1963 cited in Stliger & Shohamy， 1989 p52） proposed three components of experimental research： an experimental treatment such as a teaching method， observation or measurement of the effects of the treatment， and the randomization-random assignment of objects to groups in order to control for extraneous variables that may influence the dependent variables.

It identifies clearly causal relationships. The experimental research provides the basis for cause-and-effect relationships between two variables （Punch， 1998）. In educational situations， the experimental research is useful for teachers to test a new program or teaching method on their classes.

It is difficulty to conduct truly experimental research at class or school levels， because it is impossible to implement random assignment. Mertler & Charles （2005） list the most common threats to internal validity in experimental research： history， testing， （learning effect）， maturation， participant drop-out （attrition）， and selection bias in recruiting. To conduct a truly experimental research can minimize their influence in experimental research. However， the high validity the study has the lower external validity it moves toward. Thus the findings of it may not generalize to other contexts.

2.Quasi-experimental research

As discussed above， in real world especially in schools and classrooms， serious limitations are placed on freedom of researchers to manipulate and control many of the variables and assign subjects randomly. In this case， quasi-experimental research is used because participants cannot be randomly assigned to groups. Causal explanation can be made only when data demonstrate that plausible rival explanations are unlikely and the evidence will not be as strong as that from experimental designed. Three requirements must be met to reach a valid causal inference when using a quasi-experimental design： cause must covary with the effect， cause must precede effect， and rival hypothesis must be implausible （Johnson & Christensen， 2004）.

Because of less control， the quasi-experimental research is likely to happen under conditions closer to real educational situations. Thus， it is more likely to have external validity. Furthermore， it is very useful for teachers-conducted research and for pilot studies. It is easier to conduct quasi-experimental research because it is less intrusive and disruptive than other designs， and it is easier to get close to participants （Stliger & Shohamy， 1989）.

Because quasi-experimental researchers are neither able to assign participants randomly to groups nor can they control variables through the design. Therefore there is a clear threat to internal validity. It is possible that extraneous variables may also influence the outcomes of dependent variables. In order to remove the effects of extraneous variables and get a clear picture of relationships between independent and dependent variables， the analysis of covariance （ANCOVA） or control variable analysis may be used （Punch， 1998）.