Science method

# Design of Experiments - Science method

Explore the latest questions and answers in Design of Experiments, and find Design of Experiments experts.

Questions related to Design of Experiments

One of the most significant steps in solving multi criteria decision-making (MCDM) problems is the normalization of the decision matrix. The consideration for the normalization of the data in a judgment matrix is an essential step as it can influence the ranking list.

Is there any other normalization method for the "nominal-is-better" case besides the normalization that is possible through gray relational analysis (GRA)?

I need to understand whether the focus should be on one person at a time in such experimental designs. Kindly help me with the references if possible. Thank you.

Could any expert try to examine the new interesting methodology for multi-objective optimization?

A brand new conception of preferable probability and its evaluation were created, the book was entitled "Probability - based multi - objective optimization for material selection", and published by Springer, which opens a new way for multi-objective orthogonal experimental design, uniform experimental design, respose surface design, and robust design, etc.

It is a rational approch without personal or other subjective coefficients, and available at https://link.springer.com/book/9789811933509,

DOI: 10.1007/978-981-19-3351-6.

Best regards.

Yours

M. Zheng

Hi every one

I want to setup a new NGS based kit for my project.

In our project, other researchers used to work with one kit and It had good detection rate

some samples was False negative

The second generation kit is much more comprehensive and I want to set up it

How do you design experiment for the second kit?

Do you start with samples which were proved to be positive?

How do you choose the best minimum Input for library preparation?

This is not a laboratory experiment where we are able to control all of the variables. This is a bit more working-world application experiment. What I'm seeing in the research is a lot around the cadence of giving the spaced review before the final assessment and types of questions to use. However, as we are doing this in a workplace with optional participation, some of the additional variables we have to consider in designing the experiment are how long the tests should be available (without messing up the cadence too much) and how many times does each concept need to be retrieved in a spaced retrieval session (aka mini-testing) to be effective?

Thank you!

Hello.

I would like to ask, how should I incorporate the cytotoxicity as a response in the factorial design, since cytotoxicity depends on concentration?

My only idea is to evaluate the IC50 for each experimental unit, but it seems like a horrendous amount of work.

Are there other approaches that would be more accessible?

Thank you in advance!

Hello there!

My question concerns three cases: scientific papers, doctoral thesis and simple presentations. Which data from design of experiment are necessary in each case, which are more than welcome and which are completly redundant?

For example in full 2 level factorial design should I present half-normal probability plot, Pareto plot and ANOVA? Or perhaps ANOVA is better for papers and Shapiro Wilk normality plot better for presentations while it LOOKS better?

Another question, should I present all diagnostics plots (Box Cox, Residuals vs Factors and so on) or just mentioning that all data is correct is enough?

All suggestions are much appreciated!

Can the levels of input factors that arrive at maximum value of response (Maximum is better) considered as optimum in full factorial design ?

Hi,

I have used central compoiste design with four variables and 3 levels which gives me 31 experiements. After performing the expeirments, I found that the model is not significant. However, when I used different data (which I prevousluy obtained), I got the good model.

How do I justifiy using user-defined data? and why CCD failed to provide a significant model?

I would be really thankful for your response.

Hello.

I evaluate factors affecting the synthesis of lipid liquid crystalline nanoparticles with 2^4-1 fractional factorial design.

After the ANOVA, I have two possibilities. The first model with fewer terms (only significant in terms of p-value) has a lower R squared (0.7541). On the other hand, I can choose a model with two additional insignificant terms, one suggesting interaction between factors, and this model's R squared equals 0.8676.

What is more correct to do? To choose the model with only significant terms but with a lower R^2, or the model with more terms, including insignificant ones, but a higher R^2?

Hello!

I went through the 2 level factorial design to screen factors, and now I am moving to the characterization using CCD.

The coded alpha value in Design Expert v.13 is set to ca.-1.4. However, one of my factors assumes discrete values. I can change the alpha value to e.g. -1.5 (I go with even numbers so it works fine). My question is, how does such a change affect the design? Will I encounter any issues?

I am carrying out a multiway factorial analysis with four factors each having three levels. Two of my factors A and B have significant main effects whereas the other two C and D don't. There is no first order (AB, AC, AD, BC, BD, CD) or third order interaction (ABCD) between these variables. However, there is significant second order interaction between variable A, B and C. Can this be ignored as a higher order interaction ?

I suppose, the interaction ABC is is the difference of effect AB at different levels of C; will there be interaction ABC when there is no AB interaction?

Intuition tells me that using the mean of three measurements, which is characterized by a standard deviation already, will interfere with other statistics calculations and subsequently with fitting the model.

However I am not 100% sure which should I use. Better safe than sorry.

RBD(Randomized Block Design)

CRD (Completely Randomized Design)

Greetings, Sir/Madame

For DOE (Design of Experiments) and RSM (Response Surface Methodology), I'd like to use Python. Please post any relevant experiences, references, or Python codes in this thread.

I'm looking forward to hearing positive news from you.

Warm regards

Amir Heydari

For our Design of Experiment (DOE) we're looking for an optimal zetapotential range for a nanoparticle size between 300 and 500 nm. If possible, could you please refer to the literature about this topic with the answer?

i sudied a process using design of experiments. firstly, i used screening by fractional factorial design. results showed that 3 out of 5 affecting factors are significant. also i found significant curvature in model. so, i used RSM method (box-behnken) to better understand the process using the 3 selected factors. results showed that the linear model is the best model that fit the data. i have confused with the results. whats the reason that results from fractional factorial design show curvature but behave linear in RSM method?

Hi all,

In a experimental investigation, there are two parameters to be measured, say X1 and X2. My goal is to see how X1 varies with X2. Specifically, I am interested in classifying the graph of X1 versus X2 according to a number of characteristic graphs. Each characteristic graph corresponds to a specific state of the system which I need to determine.

The problem is with the graph of X1 vs X2 undergoing significant changes when replicating the test, thus making the classification a perplexing task. A simple approach I could think of is taking the average of these graphs, but I am not sure if this is reasonable; I am looking for a more mathematical framework.

Any comments would be appreciated.

Regards,

Armin

Any method or technique are there to maintain homogenous weed population (approx. equal proportion of weed flora, diversity and density) on entire experimental plots to check the exactly efficacy of weed control method/herbicides.

Hi

I am a physicist from Denmark planning a quantum experiment.

I need an cryostat that can go below 500 millikelvins.

I looking for any advice from researchers who have worked with cryostats before.

What are the best options for flexibility, price, maintains and operation?

Hi fellow pioneers,

I was wondering if there is a good strategy for designing a set of experiments to find the factors with the most effect on the response, which is nominal (yes/no, pass/fail type of response) instead of the typical continuous response? While for nominal response a logistic regression can be performed on available data, I doubt the usual factorial/fractional factorial design still works in this case (since they are meant for continuous response). What would be a suitable approach in this case? Kindly point me to any relevant terms/theories if anything comes to mind.

Thanks in advance.

I aim to allocate subjects to four different experimental groups by means of Permuted Block Randomization, in order to get equal group sizes.

This, according to Suresh (2011, J Hum Reprod Sci) can result in groups that are not comparable with respect to important covariates. In other words: there may be significant differences between treatments with respect to subject covaraites, e.g. age, gender, education.

I want to achieve comparable groups with respect to these covaraites. This is normally achieved with stratified randomization techniques, which itself seems to be a type of block randomization with blocks being not treatment groups, but the covariate-categories, e.g. low income and high income.

Is a combination of both approaches possible and practically feasible? If there are, e.g. 5 experimental groups and 3 covariates, each with 3 different categories, randomization that aims to achieve groups balanced wrt covariates and equal in size might be complicated.

Is it possible to perform Permuted Block Randomization to treatments for each "covariate-group", e.g. for low income, and high income groups separately, in order to achieve this goal?

Thanks in advance for answers and help.

I have a within subjects pre-test post test design experiment. The participants are mental health workers and scores of emotional intelligence are recorded before and after a night shift. I ran a paired samples t-test to compare means for pre shift and post shift scores. I would also like to explore the effects of additional variables (e.g.experience, job role and usual shift length, hour break was taken and nap length) on pre shift and post shift emotional intelligence scores. Would multiple regression be the best analysis to do this? And how would I attempt this?

Hi!

Could someone who has experience in factorial design of experiments help me with this question?

I'm completely new in this area, but I'd like to plan an experiment for initial screening to evaluate the "best" method to immobilize a protein. I want to evaluate one factor with 4 levels, and other three two-level factors (4^1 x 2^3).

But the four-level factor would be the type of reactant ("linker"), which I think probably afect the other variables (concentration, time, etc...) differently.

In this case, could I use an assymetric (4^1 x 2^3) factorial, and evaluate all the factors together, or should I "block" the "type of linker" factor, and evaluate only the two-level factors (i.e., make four two-level factorial experiments, one for each type of reactant?

Thanks in advance.

Hey guys,

I´m looking for a standard (and automatic) way to illustrate a sequence of screens used in an experiment. Is there any plataform or software that you can indicate for this purpose?

Thank you!

Hello everyone,

I need to choose a topic on design and analysis of experiment as a project. The project consists of planning, designing, conducting and analyzing an experiment, using appropriate principles and software package of design and analysis of experiments. Could you please recommend me an article or any reliable resources for the project? It must encompass 2 nuisance factors and topics such as Randomized blocks, factorial designs, 2k design and NOT RSM or CCD.

Best wishes

Design of experiment: How can we increase or change the decimal place of the terms in the regression equation generated in Minitab v19?

Is the determination of LC50 of any compound in Zebra fish during the breeding season cause any problems?

We have sufficient seeds of 120 rice genotypes. We want to evaluate the genetic parameters related to early seedling vigour through replicated trials of these 120 genotypes. We didn't want to use Augmented design. Can anyone suggest us the appropriate experimental design for the same? I shall be highly grateful.

I have two factors ( X1 and X2) and a response (Y). I want to find the values for X1 and X2 to obtain maximum Y.

I have done some tests to have a view of response in terms of X1 and X2. I fixed X2 and varied X1 (X1= 5%, 35% and 65%). I observed that Y increases with increasing X1 from 5% to 65%. I cannot increase X1 to values more than 65%.

It seems that the maximum Y happens at boundary. So, CCD design does not work. What methods of experiment design is recommended?

How can we use RSM when we have three levels?

FCC with alpha = 1?

I am working in a project to assist an experimental team in optimizing reaction conditions. The problem involves a large number of dimensions, i.e. 30+ reactants which we are trying out different concentrations to achieve the highest yield of a certain product.

I am familiar with stochastic optimization methods such as simulated annealing, genetic algorithms, which seemed like a good approach to this problem. The experimental team proposes using design of experiments (DoE), which I'm not too familiar with.

So my question is, what are the advantages/disadvantages of DoE (namely fractional factorial I believe) versus stochastic optimization methods, and are there use cases where one is preferred over the other?

Also what are the different efficient DOE and Analysis methods w.r.to machining operation with a specially sintered tool?

Experiment is done under augmented design.

Hi,

I want recommendations on how best to analyze data from my 2x2 factorial design between-subjects experiment. I have two categorical (nominal) independent variables and one categorical (ordinal) Likert-scale dependent variable. I conducted an analysis using a 2-way ANOVA because I have a large sample, and the Likert data is normally distributed. I want advice on a defensible analysis for this type of data.

Thanks!

I would like to perform a sensitivity analysis of a CFD solver. There are 8 input variables, for each of them there are 2-3 prescribed numerical values.

To evaluate one set of parameters three costly simulations (each running for 20 hours on 800 cpu cores). Budget for these simulations is limited and due to the queuing system of the HPC, it would take a long time to get the results.

I'm aware of latin hypercube hierarchical refinement methods that allows starting the sensitivity analysis with smaller budget and subsequently incorporating newer results when they're available.

But those methods works with continuous variables. Is there a method for categorical and ranked/ordinal variables?

In Design of experiments,

Response Surface Methodology,

If I conducted Experiments with variables different level example 3 factors and 3 levels, 15 runs,

Under open Atmospheric Condition,

How can I validate the model,

Because For 15 runs adopted different Atmospheric Condition,

Is there any possible solutions,

Please Suggest,

Thanking you.

Is there a Python project where a commercial FEA (finite element analysis) package is used to generate input data for a freely available optimizer, such as scipy.optimize, pymoo, pyopt, pyoptsparse?

Dear all,

I am planning to conduct an experiment for 2 IVs (categorical variable - each IV has 2 categories) and 1 mediator (continuous variable - 7-point Likert scale) on an ordinal DV (6 categories). I understand that usually mediation analysis involves regression analysis to examine the indirect and direct effect of IV --> DV and mediator --> DV, and I will be able to use the PROCESS SPSS by Hayes (2013) to estimate the moderated mediation model. However, since it is a between subject design, I am not sure if I can separate the IVs when conducting the regression analysis.

I would deeply appreciate it if anyone can recommend tests and models I can use for this study, or have any resources that I may look into to better find a suitable test. Thank you very much!

I did a product development using Design Expert for planning and evaluation. There is a response with a lot of significant effects shown in the Half Normal Plot. By testing factors A, B, C, D I got all these significant effects in addition to the interactions BCD, AB, AC, AD. Now I loose my head trying to figure out what all the interactions are about and I wondered by the way if there are commonly known mistakes leading to so many significant effects. Is it possible to get many effects, because the response might be very complex and dependent on all the factors? It would be kind of a novelty for this response.

Thanks a lot for brain-support

As a current undergraduate student majoring in Microbiology who is still learning DOE (Design of experiment) on my own, I have heard and read that an OFAT design is considered the most inefficient experimental design.

I am unsure whether that means that an OFAT design is useless to conduct or whether that means it only works for a limited number of experiments. Thank you.

Hello,

I want to realize an optimization using a Rotatable Central Composite Design, but I think it is correct to perform a Factorial Design before. This is with the aim to know the best treatment and optimize it through CCD considering new levels. Please, Could you offer an advice?

Researchers experiment with individual and group of designers. In engineering design what are pros and cons (in general) for doing experiment with two designers in a group?

Researchers experiment with individual and group of designers. In 'engineering design' what are pros and cons (in general) for doing experiment with two designers in a group? By engineering design, I mean a design problem based on mechanical, industrial design etc. By solution, I mean concept generation.

I would like to design a fractional factorial experiment that has 10 factors with 2 levels and 1 factor with 3 levels (a total of 11 factors). However, I am the only interested in 2 and 3-factor interactions involving only the factor that has 3 levels. Any recommendations on what type of design to use?

My intention is to use the design for a screening experiment.

I have four factors(F1, F2, F3, F4) with different levels( 3*2*2*3). Factors are nested each other so I am planning to do split -split-split design. please insist on the methodology for the triple split-plot design experiment.

I have data on the effectiveness of the three treatments: T1, T2 and T3 for each patient. Each variable is coded dichotomously - 0 = drug not working; 1 = drug is working. The patient could feel the effects of any of the three drugs. In such a system of variables, the Cochran's Q test seems to be the most reliable, which is the equivalent of an ANOVA with repeated measures for dichotomous (qualitative) data.

Nevertheless, design is more complicated. I am interested in the interaction with the test condition: one group of patients were given mentioned three different medications - second group in winter. So the design experiment I have is: 2 (season) x 3 (drugs) (repeated measurement) and the dependent variable is / are a dichotomous nominal variable.

Is there an interactive equivalent for the Cohran test? Technically, I could do a 2x3 ANOVA since the variable range is 0-1; however, I am looking for something more methodologically correct. Maybe just do subgroup Q tests? This also seems methodologically wrong. If anyone has heard of such a test - I will be grateful.

Dear all;

I have 4 factors to design my experiments. 3 of the factors are numerical but one is nominal.

for nominal factor I have 3 type as: A, B, C. the levels of A, B & C do not match each other.

I mean for each of them I have different levels of particle size as below:

For A - FIVE levels of particle size (F 320, F 400, F600, F800 & F 1000)

For B - FOUR levels of particle size (F 360, F 600, F800 & F 1000)

For C - SIX levels of particle size (F 120, F 360, F 400, F800, F 1000 & F 1200)

(The units of particle sizes are in FEPA grit and not important in the question)

briefly I have one factor (hardener factor) with 3 type (A, B & C) which every type has different number of levels that doesnt match each other. I need to design my experiments with these all.

could you please let me know what do you prefer me to do for designing?

Hi everyone,

I need to plan an online experiment in which each participant should watch a series of screens each containing an image (the screens/images must appear in a randomized order). During the whole series of screens, an audio file should be played, it should begin at the first image and it should end at the end of the session.

I have a Qualtrics account, but I'm not able to implement this kind of procedure. In general, as I build a page with the audio player, the audio won't be playing anymore as soon as the next screen is on. On the contrary, I need the audio to be playing in the background during the whole presentation.

Could I achieve my aim by programming something in Python / Java / Node JS / HTML? Or should I change software? Any suggestions?

thanks in advance for any help

all the best,

Alessandro

I have mostly done computational works and the transition to experimental work is slightly demotivating as I am stuck at each stage starting from whether to wear gloves for certain things to why my experiments are not reproducible!

At this stage, I am seeking an answer to how can I weigh my peptide correctly to make sure I am getting the same concentration as I planned?

For example, I want to get 1mM (just an example) concentration for my peptide and I calculated the required mass to be 0.7134 mg. Now a few questions which bother me are:

1. Since the weighing balance can only allow two decimal places, so should I round it off to 0.71 (because 3 is less than 5) or should I round it off to 0.72 (because it is likely that I will lose some peptides anyway while weighing!)

2. Even though I know theoretically the concentration of my solution, should I still measure maybe using nanodrop or some other way? (and how accurate it will be if no aromatic aa)

3. Is there anything else I should be taking care of here?

Also, if there is any authentic website or paper to read about this basics, please recommend.

Thank you all in advance.

In an split plot experiment, 4 methods of Nitrogen(N) application was assigned in main plot and 3 doses of N in subplots and replicated thrice. Along with it, in the same field, No N control was done outside the split plot design and replicated thrice. How to compare No N control with those treatments inside split plot design?

**For background:**

We have a polymer and are looking to adjust its thermal and mechanical properties to resemble the properties of a commercially available polymer to present it as a viable alternative material.

We want to do this by adding a number of additives that are known to improve those properties. We have determined a range of additives we would like to test for their effects on the polymer's properties.

**The problem:**

I am unsure of a method to determine the best possible combination/ ratio of additives to produce the closest properties. I would like to be able to test different combinations/ concentrations of additives in an efficient way and determine which combination has the best overall effect on the properties.

I initially was considering using a Taguchi Method, however as far as I understand, Taguchi Methods are not suited for properties with interaction/ variables that confound? Is this an issue for this application? Is there a method of determining the optimal combination that would be better suited?

Hello,

The aim is to evaluate the effect of silt% in cement mortar mixtures. I'm wondering which approach would be more appropriate. The mixture design or other available methods such as factorial, RSM, etc.

Thank you

For the multi-objective optimization problem is it possible to apply the concept of SN ratio to individual outputs obtained through RSM or full factorial design of experiment. Also is it possible that the design of experiment developed by full factorial can match with Taguchi orthogonal array more specifically 2 factors 3 level design problem?? Where for full factorial it is coming 9 experiments.

Thank you....

Dear all
I am planning to transiently overexpress a proten (PROTEIN A) in a HEK293FS cell line, in the same experiment I want to knockdown another protein (PROTEIN B) in the same cell line that transiently overexpress(PROTEIN A) in order to know whether knocking down of Protein B will decrease or increase Protein A.

However, the problem is that the cell line becomes confluent within 3 days, but it needs 2 days to overexpress the Protein A and 3 days to knockdown Protein B. So how should I design such experiment in a proper way?

Does anyone has experience in doing such kind of experiment?
any suggestions are welcome
Thank you.

A question just came to my mind! I appreciate any helpful answers.

Consider a problem that there are a number of input variables (for example 5 variables), and we have one output parameter. The purpose is to develop a meta-model for the problem for example using a second order RSM method. The problem is highly nonlinear, therefore using quadratic (or linear, or cubic) equations to relate the input variables to the output parameter, results in significant errors (in the whole domain). But when we subdivide the design domain (i.e input variables space) to small regions and derive specific equations for each region, the issue is resolved and the output parameter can be predicted with acceptable accuracy.

Now, my question is how one should subdivide the design region? Is there any criteria to do it with minimum subdivision of the domain? In a 2D space this can be done easily by plotting graphs and observe the graph and the points. But how can the design (input) domain be subdivided when there are more design variables (e.g. 5 design variables)?

Let's say I have the regression equation,

Y=

2 + 1.2 A – 1.3 B + 3 D + 19 B^2 – 11 AB + 13 BD

I have to attach the respective surface and contour plots for the interactive effects of A, B and D.

For both surface and contour plots,

should I do for the interactions of

choice 1> AB, AD, and BD

choice 2> AB and BD only (based on equation)

Anyone knows which choice I should go for? and why?

In Factorial Design of Experiments, each factor has different levels, one level can be considered as the base level. The cases/specimens sharing this base level can be considered as a control group. Also, randomization is similar to combination of all possible levels of all factors. In this sense, RCT and FDoE seem similar. What's your opinion?

In the response surface method, when the p-value is greater than 0.05, that component is insignificant. What does this mean? Should this component be removed from the modeled equation?

Thanks for the answers

Hello! I'm planning an experiment and I need to develop a DoE. I would like to use a software to design the experiment. My supervisor recommend me "MODDE Umetrics" , but there is not a free version available (only the trial version and I need it for a long time). Therefore, someone knows a software for DoE free and easy to use? Because it is my first time with this kind of softwares... Thank you very much in advance!

Dear Members,

I have studies one factor using different nitrogen and carbon source and even physical parameter. Presently I am looking for Statistcal optimization of media for enzyme production. with help reference, I design experiment for Placket Burman, but now confused with selection of dummy variables, which are most importance in experiment. Earlier I thought of using Nacl 0.1 % in medium as dummy variables, is it right or wrong. Are there any option other than this.

Thanks and regards

To design a proper experiment given the fNIRS signal characteristics, should one follow the fMRI experimental design recommendations (both signals present the hemodynamic delay), or are there specific recommendations for fNIRS experimental design one should care about as well? (relevant references on this?)

factors : amount of lipid with three levels

surfactant : lipid ratio with three levels

Which one is the best and versatile 'Design of Experiments' software? Can anyone suggest the list of software with the link to access the same.

Hello,

I have some categorical factors which are related to some continuous factors, In Minitab software when I use every type of design I see that factors are independent completely. for example If my salt concentration be zero, and salt type change in design, every effect of salt type is wrong, because I didn't add any of those salts at all. How can I introduce related factor as related factors to Minitab or every DOE software?

Thanks a lot

I am trying to perform a

**screening DoE**to further study influence of a range of factors for example Metakaolin content on various performances like compression strength and to further optimise my formulation in a next step.Therefore I need to set th

**e ranges for every factor which**obviously has to be done in consideration of the chemical processes. To ensure that still a geopolymerisation process will be performed after changing many factors in high amounts because of the DoE protocol my plan is to set the ranges according to what are assumed to be the mini and maximal ratios to perform geopolymerisation.It is stated in many papers and books in which direction increasing or decreasing ratios will change specific performances but I couldn't find the needed mini and maxima values.

I am very thankfull for every answer, hint or research paper you could give or recommend respectively.

I'm an master student of advanced information systems and my bachelor degree was in industrial engineering. My field of interest in industrial engineering was quality related topics like SQC, SPC, DoE, Six Sigma, etc. Now I'm looking for a proper topic for my master thesis which combines Data Science and Quality related topics.

Response surface methodology (RSM) and Multiple linear regression methods are applied to develop statistical models for catalytic reactions in order to predict conversion or selectivity within a given range of reaction conditions. Taking different process conditions, such as temperature, pressure, space velocity, time on stream as input, the statistical models are obtained. Are these methods applicable to predict conversion and selectivity by taking not only operating conditions as input parameters, but also the catalyst properties, such pore size, particle size and other properties?

If the experimental data have not been collected by DOE methods, is it always necessary to train the data for RSM by ANN, or it can be directly used to predict the model?

in D-optimal design experiment, i used 3 factors and 5 responses. for the all responses i had a lack of fit p-value less than 0.0001, what should i do? i know that mean the model is not adequate but how this affect my experiment and how can i solve such problem?

Thanks for you read this question.

I developed a pre-service teacher training module (intervention), which involves educational activities of preparation, building teams, project design(after this, teachers will practice in primary school for teaching students ), implementation, demonstration and evaluation. Meaning that the pre-service teachers will be learning knowledge in university first; after that, they will transfer their knowledge to practice in primary school. So, this module involved two phases: teachers' phase and students phase.

Now, I design two experimental research to examine the teachers' motivation (experimental one ) and students' attitude(experimental two ). The study uses the quasi-experimental non-randomized pre-test and post-test control group design. In Experiment Two, the experimental group will be trained by pre-service teacher who comes from the experimental group of Experiment One. Meanwhile, the control group will be trained by pre-service teacher who comes from the control group of Experiment One.

The question is that this experimental study design needs to use two experiments or not?

I think that this study needs two experiments because students is another subject. some say that they do not need, because this study has one intervention, and the teacher and student are the different levels.

I have to run computer simulations to find the optimal parameters for a structural problem. I have 8 factors and each of 3 levels. I can not run a full factorial design as it will take a lot of time to run 3^8 simulations. I want to reduce the number of simulations. I am confused between the Fractional Factorial method and Latin Hyper-cube method. Both can reduce the no. of experiments (simulation runs in this case), but what is fundamental difference between the two method? Which one is appropriate for this problem ?

Suggestions regarding other methods which can reduce the no. of simulation run are also welcome.

To prepare a material there are four stages where in each stage involved 3 parameters. If it was a single stage and multiple parameters I would ha e used DOE or Taguchi's technique for optimization. What do I do in the case of multiple stages ?

The Newtonian model of string vibration traces to Euler who asserted the string can have the shape of any curve that can be drawn free hand.

The Hamiltonian model traces to Brook Taylor who asserted in 1714, based on his description of harmonic motion by a pendulum as isochronous, that every point on the string passes though the center of motion at the same point in time.

These predictions are subject to experimental confirmation.

Now a days, everyone is on Euler's side, and Taylor's equations for the string are not generally known. Euler and Bernoulli attacked Taylor's version. The issue has never been settled as to who was correct. I think it is obvious Taylor was right. The wave equation has only one solution because if S is the string then dS=0. That is why it is called a stationary wave.

We can predict that an image of the string taken at two different positions on the string will show the same orbit (at dynamic equilibrium) for the Hamiltonian string but the action at two points will never synchronize in the Newtonian.

So my question is how can I form the image of a point on the string during vibration. Given the frequency of the string is maybe 440 Hz, here are my questions:

1. Do I form the image inductively or photographically?

2. How many frames a second will I need?

3. Can I make an image of the string by plugging an electric guitar into an oscilloscope?

In the Hamiltonian model points on the string only move in the z-y plane but in the Newton points move in xyz directions. Could a laser measure the length of the string during vibration?

I think the experiment I am proposing is technically possible. There must be someone that know how to do it.

There is no experimental evidence of simultaneous multi-modal string vibration, in spite of what they say on Physics Stack Exchange. Showing waves on hanging ropes or strings that are driven by an oscillator do not count.