【正文】 2. 配對(duì)試驗(yàn)設(shè)計(jì) 3. 隨機(jī)區(qū)組設(shè)計(jì)（完全、不完全） 4. 拉丁方設(shè)計(jì) 5. 反轉(zhuǎn)試驗(yàn)設(shè)計(jì) 6. 多因子試驗(yàn)設(shè)計(jì) 7. 正交試驗(yàn)設(shè)計(jì) 完全隨機(jī)化試驗(yàn)設(shè)計(jì) ? 一次試驗(yàn)只研究一個(gè)因子的作用，稱單因子試驗(yàn)，設(shè)計(jì)方法稱一元配臵法。 ? (1)首先選擇供試的 25頭豬 ? 因拉丁方設(shè)計(jì)要求處理數(shù)等于橫行數(shù)，等于豎行數(shù)，也等于重復(fù)數(shù)。如 32型，用水平數(shù)可排列成12種拉丁方。 ? 因子各水平是按試驗(yàn)期的先后有順序地反轉(zhuǎn) (交替)作用于試驗(yàn)動(dòng)物 ? 試驗(yàn)動(dòng)物頭數(shù)必須一定且兩組隨機(jī)分到相等的動(dòng)物頭數(shù) 也是反轉(zhuǎn)試驗(yàn)設(shè)計(jì)成立的條件之一 多因子試驗(yàn)設(shè)計(jì) ? 在一次試驗(yàn)中，同時(shí)研究二個(gè)或更多個(gè)因子效應(yīng)的試驗(yàn)，稱為多因子試驗(yàn) ? 多因子試驗(yàn)除內(nèi)容豐富，精確度高，取得信息量多，試驗(yàn)結(jié)論確切、具體和論據(jù)充足外，更主要的是，這種設(shè)計(jì)法，不僅可以研究多 因子各水平間 的差異，還可以分析出 各因子間的相互關(guān)系 ? 如果懷疑一個(gè)因子的作用隨著另一個(gè)因子的改變而變化，就可以通過這種設(shè)計(jì)加以檢驗(yàn) ? 這些優(yōu)點(diǎn)都是單因子試驗(yàn)設(shè)計(jì)所不及的，在畜牧試驗(yàn)中是常見而廣泛應(yīng)用的方法 ? 名稱以處理因子數(shù)而定，兩個(gè)因子的叫二元配臵法，三因子的叫三元配臵法， n個(gè)因子叫 n元配臵法 ? 因子的主效應(yīng) ? 不考慮其他因子的狀態(tài)，只考慮某一因子的水平效應(yīng)，稱之為該因子的主效應(yīng) ? 交互作用 ? 這種不同因子的不同水平間的相互影響，或者說，一種因子在另一種因子上的不同效應(yīng)，稱之為交互作用。C2D2 4個(gè)試驗(yàn)處理組。反之則應(yīng)大些 ? 例如，在育肥豬的日糧中添加某元素可大幅度提高日增重，那么試驗(yàn)規(guī)模可相對(duì)小些，反之則需規(guī)模大些 ? 實(shí)踐中，不少人錯(cuò)誤地認(rèn)為試驗(yàn)規(guī)模越大越好 ? 如在豬的飼喂試驗(yàn)中，出現(xiàn) 6個(gè)處理各設(shè) 12頭（分 2圈）共 72頭的規(guī)模，總結(jié)報(bào)告時(shí)以 12個(gè)重復(fù)進(jìn)行計(jì)算，這樣誤差自由度高達(dá) 66，是不必要且不可信的（該試驗(yàn)是將群飼的個(gè)體作為試驗(yàn)單元的） ? 另外，設(shè)計(jì)試驗(yàn)時(shí)應(yīng)注意盡量采用個(gè)處理重復(fù)數(shù)相等的設(shè)計(jì)。 Surviving the review process: an editor’ s perspective James L. Sartin Editor, Domestic Animal Endocrinology First Consideration: Should you write a paper? ? Do you have a novel and innovative hypothesis? ? Does your data provide new information to the scientific literature? ? Is your paper descriptive or mechanistic? ? Is your experimental design adequate? ? Are your statistics appropriate? ? Are your conclusions consistent with your data and starting hypothesis? Choosing a Journal ? What is the scope of your paper? ? Is the subject matter appropriate to the journal? ? Then follow the instructions to authors for this journal very closely. Choose a title ? A Role for AGRP in Appetite Regulation in a Species with Continuous Nutrient Delivery. ? The effect of AGRP on feed intake in sheep. ? Informative but attract attention Selecting authors ? Did they contribute to the conduction of the research project? ? Did they contribute to the development of the hypothesis and intellectual aspects of the project? ? Did they have a special role such as statistical analysis or a key method? ? It takes only 12 ATP to add a name, but much more if you leave off a name and insult a coworker. Abstract ? The abstract should provide the hypothesis, details of the methods and results and a short conclusion. ? It should adequately describe what you have done. Introduction ? Logical series of paragraphs to convince the readers that your paper has a solid scientific basis. ? Frame your hypothesis. ? This section is where you sell your idea. If reviewers don’t buy the argument for the project you have lost the battle to publish. Materials and Methods ? Detailed statement of what you did. ? Remember that others must be able to follow your descriptions and repeat your study. Results ? Present data in an objective manner. ? Choose the format (tables, figures or micrographs) that best illustrates your for big data sets. ? Use adequate figure legends so the figures can be understood alone (may differ with different journals). ? Identify important trends. ? Do not interpret your data. That is a discussion. Discussion ? Tell your reader what your data mean. ? You should indicate how your data answers the hypothesis—from Introduction. ? You should talk about how your data relates to the scientific literature. ? How has your data advanced scientific knowledge. ? Avoid too much speculation. ? A concise summary or conclusion in useful. References ? Your references should cite relevant articles in the field. ? Do not use more than needed – check a current journal to suggest a number. ? Check the journal style. ? Make sure the information you cite is accurate. ? Use recent references. Submitting your manuscript Authors ? Cover letter ? Make sure all authors have read and approved the papers submission. ? Some journals require that all authors sign a form. Suggest reviewers ? Many journals offer the author the chance to suggest reviewers (provide plete contact information including phone, FAX and ). ? Most journals will allow authors to specify reviewers that they do not want to see their paper. ? Some editors look at your references list or search PubMed using your keywords for reviewer names. Common reviewer plaints ? Paper doesn’t add anything new to the scientific literature. ? This paper is fine but is just descriptive. ? For some journals—this has already been done in the rat. ? Results don’t prove the stated hypothesis. ? Did not follow instructions to authors. ? Did not follow animal welfare regulations or concerns. ? No hypothesis ? Figure legends do not provide sufficient information ? Statistics are not appropriate for the experimental design. ? Methods are not adequate. ? Poor quality illustrations. ? Disanized progression of the paper. Primary problem for Chinese authors? ? Chenglish. ? The paper cannot be evaluated due to improper English. Reviewers may refuse to evaluate paper. ? A good paper scientifically is rejected only on the basis of language. ? International Science editing ? Asian Science Editing Responding to the reviewers ? Do not respond antagonistically. ? Carefully consider and respond pletely to all reviewer ments. ? Identify by line numbers and page numbers where all changes have been made in the manuscript. ? Do not send a paper back with no changes. ? Failure to follow a reviewers advice can damage your chance of publishing. Don’t make them angry. Rejection ? Most choose another journal and try again. A rejection may only mean the reviewers didn’t like your paper or you chose a wrong journal. ? Endocrinology only accepts the top 25% of papers. ? Lower scientific impact journals may publish