Before I can give an answer, I need additional information:

1) have the concentrations been measured/applied within the different plant parts (i.e. different measurement points) or has there been for example a root with 2 g/L and another root with 4g/L (sorry I am not an biologist, don't know how you do this)

2) Do you want to know if there is a difference between the different plant parts (e.g. root compared to leaves)?

Apparently, your germination index is your dependent variable, but depending on the answers to my questions, you have to use different statistical models.

independent variables are the input which you varied or called variables

dependent variable means response or end point or result on changing dependent variables.

I agree with Arshad. The independent variable is one that you manipulate (concentration of extracts). The dependent variable is one that will respond to the change of independent variables (germination).

EXAMPLE SUGARCANE,

CANE YIELD IS DEPENDENT VARIABLE AND CANE WEIGHT AND NUMBER OF MILLABLE CANES (STALKS) ARE INDEPENDENT VARIABLES. BECAUSE CANE YIELD IS OUT COME OF THESE TWO TRAITS. DEPENDENT VARIABLE IS MOSTLY ONE AND INDEPENDENT VARIABLES MAY BE 20 TO 50, DEPENDING ON NATURE OF STUDIES.

CANE YIELD = CANE WEIGHT X NUMBER OF  MILLABLE CANES

CANE YIELD 120 TONS = 1.20 KG CANE WEIGHT X 100,000 MILLABLE CANES

As you wrote you have two independent variables, the first is the plant parts and the second is the concentration. The dependent variable is germination parameters of wheat. However, there is special technique for data entering in order to run two-ways ANOVA.   

Thanks all of you my colleagues, your depth of knowledge and illustration with examples really diminished my confusion; now i am clear that plant parts and their concentrations are independent variables while germination parameters are dependent. thanx lot

Since the objective is to test the  effect of different plant parts (roots, leave, stem)  on germination parameters of wheat, your linear model should be:

Y = mu + Ri + Lj + Sk + interaction terms + error

where Ri = effect of Root, etc and Y = (one of the )germination parameters.

Since you are conducting a physical experiment in which you control R, L, S, these are your Independent variables, and germination parameters are your Dependent or Response variables. 

as i understood from attached file, you are evaluating Contimous as well as catagorical input variables(Independent variables). Contimous variables are concentration of root, stem peel and leaves and catagorical variables are root, stem peel and leaves. Your  response variable (dependent variable) is germination of seed

dependent variables= germination parameters of wheat.

independent variables = 2 (categorical =1, continuous = 1)

Categorical = plant parts (roots, leave, stem)

Continuous = concentrations (0, 5, 10, 15, 20 g/L)

Dear Varinder and Asaimonat,

I disagree with your explanation. Why is concentration a continous variable? It has discrete categories 0,5,10,15, and 20 g/L. Although they are equidistant doesn't it mean that this is continous! You may ask if this categories represent all possible categories, which leads to the question if you should use a random or fixed effects statistical model.

I am sorry but I disagree with Rainer, concentration is a continuous predictor here. 

Just follow the chornonolgical order, which one occures first will be the IV.

I understand you have two independent variables Parts of the plant and the concentration of the aqueous extract veriable is dependent on the effect or response of these two factors is: Parámateros germination of wheat

I understand you have two independent variables Parts of the plant and the concentration of aqueous extract. The dependent variable is on the effect or response of these two factors is: Parameters wheat germination

Dear Prof Singh,

if you disagree, how would you treat the following variable:

Imagine you test the effekt of a pill with any drug and you have three (or four or five) independent groups. One gets 0 pills (waiting group), the second 1 pill and the third group 2 pills. Dependent variable could be for example reaction time or whatever. How would you enter this into a regression model, predicting reaction time? As a metric single predictor or as two dummy coded variables with the control group as 0 (could also be any other coding)?

In my opinion it makes more sense to use the dummy coding and treat it as discrete categories instead of a metric variable. For a better generalization a random effect model could be more appropriate.

But please share your thoughts on this.

The values of the 'dependent' variable DEPEND ON the fixed, chosen or selected values for the 'independent' variables. Because these two terms appear confusing other names are used like 'response variable' (in place of dependent variable) and 'predictor variables (in place of independent variables).

Some of the above answers have introduced other aspects of variable classifications that are important but do not specifically target the question. For example, the confusion about continuous and categorical variables: Concentration is a 'continuous variable' even when fixed levels of concentrations have been used. One can pretend that it is a categorical variable, for purposes of analysis, but you will lose information unless you view the results against the original metric scale.  For example if doses were given in measurements of  'numbers of pills' , you have immediately lost information on the content (actual dose concentration) of those pills. 

Understanding variables  and how they are classified (including different types of metric scales for continuous variables) is fundamental to understanding statistics. Some variables can appear to have overlapping (non unique) classifications. However the key to success with statistics is in 'understanding' variables, how they might be classified, and how that classification might be used in analysis; not rote-learning about what category a variable belongs to.

For Abdul Majeed  questiion on the statiscal analyses of his experiment and data given  as attached file I could not see the design of experiment and replicates. The data collected  are not clear  either germination or concentration of extractions.  From the title  of  experiemnt  I could see the extract  concentration from various parts  of Nigella.   I dont know which concentrations of 0,5,10,15 and 20 g/L while in the file I see this

Control 0  2.0  4.0 6.0 8.0

I would suggest to  verify  the design of experiment  and replicates/number of observation used and which actual  independent variables and dependent variables. To analyse by experimental design we need replicates  or numbers of observation  for SPSS or SAS to analyse the data.

Dependent variables - observation or  data taken as effected by the factors

Independent variables - factors studied  or treatment  variables.

If you want to analyse two way anova by SPSS the data should be as attached file if   parts of plant and concenctrations are the factors being studied while %germination is data of observation or effect of the factors. Assuming you have 4 replicates.

You need to take a basic stat course. Actually, most stat programs have books you can buy that use the program to learn basic stats. I highly recommend it.

One way to look at it is to think about the independent variables as the input of a process and the dependent variable as the output. You can vary the input, and observe the effect on the output. The process that does this is the 'thing' in the middle that manipulates the input (independent variables) to get the output (dependent variable). The middle bit can be for example a linear regression model, a logistic regression model, a neural network etc. The various models that can be used are taught in statistics and machine learning.

I agree with Iqbal, I just would like to add that there are situations in which the distinction is not so clear. For example, in gaming, one collects data on X1=Slots revenue and X2=Buffet revenue. In this example, one could predict X1 as a function of X2, or X2 as a function of X1.

The germination parameters of wheat are dependent on the concentration. Therefore the germination parameters of wheat are the dependent variable. the independent variable is the one you can change according to your requirement which is concentration. Hope this is useful.

The variable the researcher measures as output are called dependent variables in this case your dependent variables are your concentration -0,5,10,15 and 20g/l while your independent variable are variable you do not have control over and in this your study, your independent variables are  Roots, Leave and Stem . Hoping this will be  helpful.