Subscribe to:
Post Comments (Atom)
Followers
Blog Archive
-
▼
2009
(18)
-
▼
August
(18)
- Discussion Card for Creswell's Ch. 1
- Discussion Card for Creswell's Ch. 2
- Discussion Card for Creswell's Ch. 3
- Discussion Card for Creswell's Ch. 4
- Discussion Card for Creswell's Ch. 5
- Discussion Card for Creswell's Ch. 6
- Discussion Card for Creswell's Ch. 7
- Discussion Card for Creswell's Ch. 8
- Discussion Card for Creswell's Ch. 9
- Discussion Card for Creswell's Ch. 10
- Discussion Card for Creswell's Ch. 11
- Discussion Card for Creswell's Ch. 12
- Discussion Card for Creswell's Ch. 13
- Discussion Card for Creswell's Ch. 14
- Discussion Card for Creswell's Ch. 15
- Discussion Card for Creswell's Ch. 16
- Discussion Card for Creswell's Ch. 17
- Discussion Card for Creswell's Ch. 18
-
▼
August
(18)
Quote: (p. 183) "A good rule to follow, however, is that the more positive the response or the higher or more advanced the categories of information, the higher the assigned number."
ReplyDeleteFact: Statistics is the calculations of values based on numbers. I cannot believe that this is the first time I have seen a definition for statistics. I have been intimidated all this time by something that I did not even know the definition of. Nonetheless, examples on pages 192 and 193 make my head want to explode! So, I am still scared, just an informed coward!
Question: If you plot the means of many samples, no matter what, a normal curve would result. How can this be? Maybe we can request the assistance of a certain mathematics professor. ;)
Corey Nickell
ReplyDeleteQuestion: I got a bit lost in the Hypothesis testing section. Specifically Null Hypothesis. Are we basically trying to set a hypothesis where there is a difference or there isn’t?
Quote: Pg 190, “You can substitute numbers for missing data in a database for individuals.” This surprises me since I figured that “good” research wouldn’t have any sort of replacement going on in the data.
Interesting fact: I am always happy we I see something familiar in this book, it makes research seem less intimidating. That brings me to the figures! Presenting the information ind a chart, graph, or picture is so teachery!
Tanya...lol@the mathematics professor!!! He is being fed so at the very least he could help us out!
ReplyDeleteQuestion: How frequently do researchers encounter "missing data" when interpreting data? Is it the researcher's fault if participants do not respond or provide incomplete data? On page 189, Creswell says "When individuals do not respond, something is wrong with your data collection, which may indicate faulty planning in your design." How can the researcher control how or if participants will respond?
ReplyDeleteQuote: pg. 209, "It is important in concluding a study to advance limitations to the research, noting potential weaknesses that might have affected the results."
Fact: It surprises me that in presenting results a researcher my suggest future research questions. Two thoughts; wouldn't a researcher want to use his ideas in his own future research, and it seems a bit presumptious to assume others would utilize your suggestions for future research.
Interesting Fact: "Before conducting an analysis of scores, researchers consider what types of scores to use from their instruments. This is important because the type of score will affect how you enter data into the computer file for analysis." Page 184
ReplyDeleteQuote: "How should you handle missing data? The most obvious apporach is to have a good instrument that individuals want to complete and are capable of answering so that missing data will not occur." Page 189
Question: When reporting results, should you represent data using tables or figures or a mixture of both? Is there one preferred way?
Corey, your quote surprised me too when I was reading the text. I would have thought the researcher would just omit that one. I guess not necessarily.
ReplyDeleteMrs. Dalrymple and Corey,
ReplyDeleteI got a little overwhelmed in parts of this chapter, as well. The hypothesis testing section, specifically about the significance levels and "p" and "df".
Interesting Fact: Mean is average, mode is most occurring, median is the middle # (what I tell my math students is the middle of the road), range is the difference between the largest and smallest...Yes, something I know! :)
ReplyDeleteQuote: (p185) "...responses to single questions may not be reliable and may not accurately reflect an individual's score." This may be due to the participant not understanding the question, or possibly question biases. Therefore summed scores are often used.
Question: If one study intentionally replicates another but does not have the same limitations, can the researcher say that their findings refute or support the findings of the original research or are they too different to compare?
Question: Is one way to score data better and more effective than some of the others? One more relied on than others?
ReplyDeleteInteresting Fact: Page 190 - You can substitute a number for missing data in the data base for individuals. (I don't believe this would provide the researcher with accurate information. Won't this scew the results and make the results inaccurate?)
Quote: Page 200 - "Although me might have anticipated that the "Punkers" group had more smokers than the "athletes" group had more nonsmokers than smokers, our statistical test didn't find these results to be present." - Interesting! What were the results?
Dana - Amen to the confused and overwhelmed feeling! I went back to reread the hypotheses section several times to make sure if I was thinking and understanding correctly. I need some clarification on this. Just read through the notes I took and I am shaking my head in not understanding.
ReplyDeleteJust found a better quote!
ReplyDeleteQuote: Page 202 - "In educational research, we need to be cautious about saying "there is a difference" when one actually doesn't exist."
* "The power of quantitative hypothesis testing is the probability of correctly rejecting a false null hypotheses."
Dana - You mentioned responses to single questions not being used. I have run into this with benchmark assessments. We are testing students on a concept(CA Benchmark) and scoring of they have mastered the concept based on one or two questions. This to me is not an accurate portrayal of there ability to have mastered the skills. The more questions the better evidence of understanding or misunderstanding.
ReplyDeleteChristina - I just looked at my notes to see what I wrote about a preference of using tables and/or figures in reporting results. In my notes I wrote down that tables are preferred. What I am gathering from this is possible because there is more room on a table for more detailed data? And less interpretation of the results than if I chart, graph or a picture would be used.
ReplyDeleteAnn,
ReplyDeleteGreat question! Your research question really dictates what method you should use and what you are trying to improve. Then, you choose what method you will use and how you will score etc. What is your goal with those scores? What type of analysis do you need? Then, you choose which one will work best for you.
I covered this question in class twice last night. Make sure you watch the video about this one--the missing data quote.
Keep up the good work,
Dr. Hendrix
Question: How do you decide if you should use summed scores rather than single-item scores, or should both always be included?
ReplyDeleteQuote: "In reporting research today, it is useful to report all three estimates of your population: hypothesis testing, the confidence interval, and the effect size." page 196
Interesting Fact: Even though tables and figures summarize information from statistical tests, a researcher also needs to describe in detail the results from those tests.
Interesting Fact; When individuals do not respond, something is wrong with your data collection, which may indicate faulty planning in your design.( pg 189) I would not want to get to this part of my research!
ReplyDeleteQuestion. Should I be worried if I have to use inferential statistics to analyze data?" I know about not deep enough inferencing!
Quote:"When you study the sample and obtain scores, several approaches exist for determining if the sample scores you recieve are a good estimate of the population scores." pg 195
Ann G.,
ReplyDeleteYou will look at your research questions and what you are trying to find, then that will direct you towards whichever quantitative route to go in then.
Take care,
Dr. Hendrix
Shelly,
ReplyDeleteDo not forget your question!
Take care,
Dr. Hendrix
Quote:
ReplyDelete"The power in quantitative hypothesis testing is the probability of correctly rejecting a false null hypothesis. The resarcher can fail to reject the null hypothesis when it should not be rejected because there was no effect." (2005) p.193
Fact: Type II erros occurs when the null hypotheis is rejected by tht e researcher when it is actually true. This could be confusing for an amateur researcher, but not knowing what the results show and how to interpret them accurately can give false information and lead people to believe the facts of a study were true when they were not.
Question:
I wonder which type of errors occur most frequently in educational research... type I or type II?
This comment has been removed by the author.
ReplyDelete