Manual for doing an educational research
Информация - Педагогика
Другие материалы по предмету Педагогика
ividuals result in a multiple item. Usually raw scores do not give enough information. If the students result is 85 out of 110, it is difficult to say if it is good or bad. If it is the highest score in the class, it will be good; or it can be average; or even the worst. That is why derived scores are used more often. In this measurement way the data is interpreted in a comparison to other results.
Age and grade equivalents show already measured students abilities at certain levels. These data show the average age and level of knowledge students can have in different grades. Usually this equivalent is applicable for tests results. For example a researcher wants to measure students reading abilities. It is already known that in general people read about 170-200 words per minute, but this average level is not applicable for school students. That is why a researcher must take into consideration participants age and grade information in order to make objective conclusions.
Rank scores show the position one student or a person takes in comparison to others who have the same experience. Ranking list can be used at schools where all the school-leavers will be ranked according to their grades from the highest to the lowest. The perception of ranking list can vary according to the scale. For example if there are 50 applicants for a position and they are ranked on the scale from 0 till 5, there will be no difference between applicants sharing the same rank, but if the same applicants are ranked from 0 till 50, then much more information will be takes into consideration and the difference within 5 or 10 points can be crucial.of central tendency in descriptive statistics are used to calculate the average among all the scores. Within these measures it is possible to identify further calculations. Mean is measured by dividing the sum of all possible scores into the score group of students got. Median is what can be found in the middle of score distribution, while mode is the most common amount of scores students get. If the group of students presents on the average level without a big difference of knowledge level, then the median and the mode can be equal. But if there are mane extreme points, then these data will be different. Standard deviation is the most used calculation in statistic data. There is a special formula to calculate it, but it is based on the mean score.after the data collection process a researcher can find out that some data are missing. Missing data means that a researcher planned to get special amount of data from different sources (participants, documents or other information) but either because of carelessness or some unpredictable situations the plan did not work. If the expected participants are students, it can happen that some of them do not come to school because of sickness, there is always a human unpredictadness. In this case if many participants are missing, probably it will be better to postpone the data collection process for few days. It is always better if everything works according to the plan, but if some data are missing, then the idea of the whole research and later results can be useless. But even if some data are missing, the researcher must be honest and write it in the study, but never try to cheat by giving wrong information.analysis process can be always conducted on two levels: multilevel or one level grouping. For example a researcher collected data about students perception of teachers work with seven-graders. On the one level grouping a researcher will concentrate only on these particular students and on one or several chosen teachers. But if the data are analyzed on a multilevel then the researcher has to look at the situation in a wider way and look at teachers work on the whole school level.statistical data is also a long and quite complicated process. Usually in quantitative research studies there is a huge amount of information which needs to be analyzed. More and more often it is done with the help of the computers. Nowadays there are several computer programmes made for data analysis, such as SPSS (Statistical Package for the Social Science) or SAS (Statistical Analysis System) and others. The usage of these programmes is not very easy and it is better to have some experience in working with them. If a researcher is inexperienced, it can take a long time of him to get used to it; that is why some students prefer to search for help by hiring computer consultant. Data analysis process is one of the final stages and it is recommended to check the data for accuracy before submitting it. The data must be checked very carefully, from the very beginning of typing data, and then different formulas for calculation and results.the data analysis is finished a researcher should take care of the study and keep it a safe place. It is recommended to have several copies of it and also still save the raw material just in case if later he or she decides to wants to have a look at it or show it for the committee. But it is not common to re-analyze the data, it is better not to make any correction after the process is completed.
Selecting a sample
The selection of participants for the research must be well thought-out. The researcher must be very attentive and be able to give good reasons proving his or her decision. The number of participant can be different in accordance to the research scope. If a researcher wants to study the influence of the national curriculum on schools performance, then probably several schools in a municipality or even several municipalities will be involved in the data collection process. But if a researcher is interested in some particular cases, such as life of some celebrities, then even one participant will be enough. One of the criteria for judging research results is population validity; it means that results concluded from the research study can be applicable for the population of the city, region or even state. In order to achieve the real validity, the number if the participants must be sufficient. It is worth remembering that these participants must be chosen randomly, without any privilege. A different way of sampling is a purposeful one. The topic if this study has to sound very narrow, such as Challenges graduate teachers face during their first year. The idea of this research is to know some peculiarities, go into details, and as a result of this the participants will be chosen very carefully, only those who suit the requirements. One more way of arranging a sample is replication logic. It means that a researcher will base the study to check or to prove one of the previous studies. An example of this can be a teachers leadership style. Suppose that according to one of the studies a biology teacher was successful by using a particular leadership style. The next researcher can use this study as a foundation for his or her study. And finally there will be several studies proving that biology teachers in a particular situations or cases can be successful by using a special leadership style. The biggest quantitative research which can be done is the one where the whole population will be covered. But practically it is hardly possible. That is why there are two common ways: to identify a target population or accessible population. In the target special group of people can be involved, such as all biology teachers in New Jersey, or special teachers in Tampere. The idea of this target group that it sounds not very huge, but the outcome can be applicable for the same group of people but wider. Another way is accessible population, the research done in this way can cover larger group of people, such as all school-leavers in Kuopio. Of course practically it is not possible to involve all the school-leavers, but a researcher has to choose those schools which are close the place where he or she lives and some neigbouring. Then some generalization must be done. First the results from the participants who were involved will be generalized to the accessible population and then it can be generalized to the target group, this process is called inferential leaps. And these generalizations can be done much easier if participants were chosen randomly, if everybody was equal and could participate. But if there was a selection of people for the research, then the researcher has to prove that any generalization is possible, to show that some wordsities between participants and accessible population and then target population can be found. Choosing participants for the sample randomly is more challenging than a systematic way. Random selection takes more time, it can be done either through a table, where each possible participant has it is own number or with a help of computer programme. While in the systematic way more logic can be found, such as every tenth or every 37th student or a teacher.idea of choosing a sample in qualitative research is different from a quantitative one, the most common sampling is a purposeful one, within which Patton identifies 15 strategies.
.Extreme or deviant case sampling is concentrated on unusual or strange cases. These example can give some understanding of the typical cases. But the main problem of these studies is that researchers can dismiss some findings.
.Intensity sampling happens when a researcher chooses the best representatives of the field. They can be the principals of the most successful school or the most famous teachers. Some people can be skeptical about it, because it does not look like their case, but in general this sample will give deepen knowledge and understanding of the situation and maybe even inspire some followers.
.Typical case sampling involves usual people without any extremes. Usually a researcher will use this sample when a new theory or approach needs to be test