To: Methodology of medical scientific research saves all take an examination ofing to training class student the 8th part

THC Kills Glioma Cancer Cells - Medical Miracles from Europe

26, how does illustrate apply Latin to just be designed in animal experiment

Every 3 elements experiment, the level that is like every element can be accomplished equal when, and when interact is not being had between the element, all can use Latin to just be designed. In animal experiment because the condition is opposite easy control, latin square design is having extensive use, and the order in animal experiment is influential to experimental result because of usually, must use Latin to just be designed at this moment.

For example, use Latin to just design pair of garden burnet charcoal, cattail pollen pink, hairyvein agrimony 3 kinds of Chinese traditional medical undertake the hurried outside small mouse system coagulates action undertakes comparative, take 16 small white rat to divide 4 groups randomly, give celiac inject the fluid of water simmer in water of 3 kinds of Chinese traditional medical and physiological saline respectively, experimental design sees next tables (Experimental data is surely inside bracket, A- garden burnetB- hairyvein agrimonyC- Pu HuangD-NS)

Rat date

Ordinal

∑ X

One

2

3

4

1

A ()

B ()

C ()

D ()

2

B ()

C ()

D ()

A ()

3

C ()

D ()

A ()

B ()

4

D ()

A ()

B ()

C ()

27, does cross section investigate commonly used where two kinds of methods? Respective use condition how?

Commonly used method has census and sampling investigation two kinds.

Developing the condition with essential census is: (1) the labor power that has Zuo , material resources and equipment are used at discovering case of illness and seasonable cure. (2) the disease sicken rate of institute reconnaissance survey is higher. (3) the technology of operation of law of check proved recipe of the disease is very not sophisticated.

The requirement that sampling investigation has is: (1) research target is had representative, the sample size with sufficient demand and avoid phyletic deflective to intervene. (2) had not fitted old stuff. (3) the situation that does not suit to need census general to treat (4) do not agree with the disease with sicken inferior rate.

28, which a few kinds does common random sampling method have? What characteristic do each have?

Common random sampling method has pure and random sample, the system samples randomly, statified and random sample with cluster randomly, its are respective the characteristic is:

(one) pure and random sample: Also say to sample simply randomly, fromNThe method is used in the objectN, make an example. It is the simplest, the basiccest sampling method, its are important the principle is overall the probability that each medium object is smoked is equal. The advantage that samples simply randomly is easy understanding, the result of example can project overall in go, its error of sampling is calculated more easily also. But sample simply to also have a few limitation randomly, should construct for example a sampling draw a frame round that can offer draw-out and simple random example often very difficult, executive likelihood is more troublesome also.

(2) the system samples randomly: Be according to certain order, mechanically every other a certain number of units, draw-out the sampling method of an unit. The first example uses the method that samples simply randomly draw-out. The advantage that the system samples is: (1) do not need to know beforehand overall the odd digit inside, (2) in crowd spot Yi Jin goes, (3) example is from distributing draw-out in overall unit, distributing evener, representative good. Main drawback is: If overall in of each units distributing have periodic tendency, and draw-out removed just is its cycle or its multiple, make example generation slants possibly quality.

(3) statified sample: Statified sampling is a kind of two phase actually sample, be first will overall do not overlap into a certain number of each other by certain and characteristic cent child overall (or layer) , be in next each stature is overall (or layer) in according to random method draw-out element, it is to use simple and random sampling technology normally. Statified sampling characteristic depends on: (1) it is all and important to can assure child overall represent somewhat in example, make example has more representative; (2) error of sampling is smaller, have the quality of the simple palpability that samples simply randomly, precision prep above samples simply randomly, carry out should facilitate more than sampling simply randomly. Because this is statified,sampling is very commonly used technology.

(4) cluster sampling: Be do not overlap overall cent for a certain number of each other first child overall (group) , be in next all child overall in, random ground is draw-out one part, right in smoking these all group of unit inside undertake investigating. In cluster sampling, draw-out is merely child overall (group) an example, and be in statified in sampling, all child overall (layer) by draw-out as the foundation that samples further. Advantage, that is organizing the job to go up to go to the lavatory quite namely, need only about group sampling casing and need not group inside the list of sub unit. But as a result of sampling unit too concentration, because this counts same situation to fall in example, the precision of cluster sampling is general under sample simply randomly. To obtain with sample simply randomly same precision, the sample size of cluster sampling (to sub unit character) should be simple and random sample a certain number of times. This multiple calls design effect, general design effect often takes 2 right-and-left costs.

29, what does the main bias in cross section research have? How to undertake controlling?

Common bias: (one) choose bias, namely example cannot be represented overall, include: (1) alternative bias (2) without respondent bias (3) survival bias

(2) information bias, the error that appears when obtaining information namely, include: (1) memory bias or report bias (2) investigation personnel bias (3) measure bias

Quality controls a method: (1) randomization principle, undertake those who study a target strictly choose according to sampling design program (2) the comply with sex that raises research target and rate getting check, seasonable analysis does not have respondent reason, filling leakage is investigated (3) choose excellent instrument facilities to undertake corrector beforehand, those who assure to check a result is accurate with reliable (4) training strictly the member that investigate makes supervisory be controllinged with quality unity investigate program and method to its (5) the check that makes good material, check the job (6) the statistic with right choice analyses a method, notice differentiate and analyse is jumbly element and its influence.