Behavioral Science Statistics Statistics for the behavioral Sciences
The third part ~ The deduction of the difference between the mean and the average
The deduction of the difference between the mean and the average is a total of eight chapters, all of which are statistical methods. Each method uses the average sample as the basis for inferring the overall average .
Is that the average number of samples can be inferred to the overall average of various methods ~
The 8th Chapter hypothesis test
- Foreword: Is it accidental, or is there a real difference? This is one of the purposes of inferential statistics, hypothesis testing.
- You read this passage '
- Hypothesis testing helps researchers to differentiate between real and random patterns in data. In the study, the aim was to determine whether the results showed a true relationship between the two variables, or that the relationship was caused by occasional random fluctuations. ~
- 8.1 Logic of hypothesis testing
- It is not always possible or impossible for researchers to observe every individual researcher in the population.
- Hypothesis testing is a statistical method that uses sample data data to evaluate assumptions about the overall parameters.
- Told four classic steps, ' one by one ...
- 1?? void hypothesis (H0): Handling without effect H0: u0 = u1 is also called 0 hypothesis .... In addition, it is called the alternative hypothesis H1, which means that there is a change, difference or correlation between the general population.
- 2?? Setting the judging criteria
- 3?? Collect data and Calculate sample statistics
- 4?? Make a judgment
- 8.2 Uncertainty and error in hypothesis testing
- First Class Error: there is actually no effect, but the data will instruct you to reject the null hypothesis, suggesting that you have an effect. Do you know the probability of the first type of error is the probability of a? Because the significance level A is born for this oh ... = =
- As long as the researcher rejects the null hypothesis, there is the risk of the first type of error. Similarly, as long as the researcher cannot reject the null hypothesis, there is a risk of a second kind of error ...
- The second kind of error: the opposite of the first class is so simple ah ... There is really a significant effect, but the data is not detected ....... It's so simple ...
-
- So the son said ... A look at this form is actually quite simple.
-
|   |
Real case |
| no effect ho is correct when |
effect exists, ho is wrong when |
| The experimenter's conclusion |
deny ho |
First class error |
conclusion correct |
| accept ho |
conclusion correct D |
second class error |
- 8.3 Examples of hypothesis tests
- In statistical tests, significant results mean that the void hypothesis is rejected ~!!! ... spss are used with P-values, however our usual calculations are with Z-fractions.
- Well, actually understand the statistics behind the principle, when running SPSS will not faint like a cabbage ...
- Hypothesis of the Z-score hypothesis test ... There are three prerequisites, pay attention to Oh ~. It is a random sample, independent observation, processing will not change a value . This last point is that we assume that the standard deviation of the unknown population (after processing) is the same as the population before processing. I don't know what that means. The last one is a normal sample distribution .
- 8.4 Directional (Single-tailed) hypothesis testing
- Does it mean that the assumption at the outset is expected to increase or decrease the effect? We don't know ... So
- Suddenly remembered ... I used to Envy B District Two comprehensive front of the dormitory students can always go to two books. In fact, I lived three or 3 years, through the playground is the library .... = = Is this not the same effect, the mood is not good, you can go back to the bath, and then come back to read. However, there is no two in front of that hostel so close ... Anyway, reading is good ~ after still hope to read ~
- This is why directional testing is generally referred to as a single-tailed test!
- The difference between the directional test and the general hypothesis test is that the first step and the second step ... Other same ~
- When the first analysis results in the use of double-tailed tests that do not produce significant results, you should never use a single-tailed test as a remedy to try to get a significant result of the study. ~
- 8.5 about hypothesis Testing: measuring effect size
- Cohen D-value = mean difference/standard deviation ...... At least know Cohen D value!!
- .......... This book, 220 pages, has the size of the Cohen D-value evaluation effect.
- 8.6 Statistical effectiveness
- Measuring the effectiveness of statistical tests ...
- A little bit of a feeling. Statistical efficiency is the probability that a null hypothesis can be correctly rejected, that is, the probability that the effectiveness is to be recognized as a real processing effect.
- If you are commenting to get someone to answer you ... Or expecting someone to answer you ... Better not reply. ... It's a rare thing.
- The effectiveness of the white is not to check if there is no effect.
- Three factors affect the size of the performance:
- A level increases, the effectiveness will increase.
- The effectiveness of single-tail test is higher than that of double-tail test.
- The efficiency of a large sample is greater than that of a small sample.
Oh, my God... This chapter has finally been read .... Oh, my God! ~ ~
* Summary written in front of it, no disc no progress.
- The most important ... One... There's a formula and vocabulary behind it ~
- z=m-u/aM to Chinese z= sample Average- assumed overall mean/standard mistaken for this hypothesis ... This word is too critical, in fact, from the first step is assumed!! ~ ~
- Then this supplement is God's knife! The Z-score test is rarely used in practical studies, and the problem with Z-score testing is that it needs to know the overall standard deviation! ~, and this information is usually unknown! So there will be thousands of statistical methods behind ...
- Time goes fast, not to add and practice, it is difficult to catch up with knowledge ...
March 17, 2016 Thursday inferential statistics come on, ~-. /The actual finishing in 2016/12/18..2016 year is fast, this is a magical year ~ Brexit, Trump became president. I can also enter the academy, the student career is almost gone ... All of this is going forward, only. We continue to keep learning and tidy mentality can not be broken, exercise and love family action can not stop ...
Recently Love self-study ~ Continue ~
Behavioral Science Statistics Chapter 8th
