# Statistics: refers to collections of quantitative information, and methods of handling data,

HideShow resource information
• Created by: Cyn:)
• Created on: 19-02-14 22:55
• Introduction to research and analytical skills and experimental design
• Dependent vs Independent
• Independent variable: this is the factor that can be changed.
• Dependent variable: The factor that consequently changes.
• Experiment is an investigation of the relationship between two things by deliberately producing a change in the other.
• Two types of experiments
• 1) OBSERVATIONAL
• Experimental conditionsare NOT manipulated
• One or more responses measured and correlated with one or more 'environmental' measurements.
• This type of experiment is realistic as it is 'natural' and not manipulated. A good general experiment unlike manipulative experiments.
• 2) MANIPULATIVE
• Experimental conditions are manipulated and controlled to investigate cause and effect relationship.
• This approach allows relationships to be deduced.
• Manipulative experiment has excelled control of independent variables and good deduction ability unlike observation experiments.
• Research hypothesis
• Null hypothesis (Ho)- represents a theory that has been put forward 1) because it believed to be true or; 2) because it is used as a bsis for argument but has not been proved.
• Research Hypothesis
• Null Hypothesis
• This hypothesis assumes the two sample means drawn from two populations are the same. Even though the means are different
• Alternative hypothesis
• This hypothesises simply the differences.
• Repititions
• How many observations need to be made to ensure that the results collected are valid.
• eg: sample size
• The term replication means that a sample should normally consist of more than one sampling unit
• The greater the variation in data, more data required to produce a value in confidence.
• The reliability of a sample average depends on the variation of data.
• Randomisation
• The process by which experimental units are allocated to treatments: by random process therefore no biased approach.
• The treatments should be allocated in such a way that each treatment is equallylikely to be applied to each unit.
• Randomised block design-this can be used as random sampling cannot be easily achieved. It isn't possible to sample a whole population evenly.
• 1: Divide the population into different groups.2: take sample from all the different groups.
• Two types of experiments
• 1) OBSERVATIONAL
• Experimental conditionsare NOT manipulated
• One or more responses measured and correlated with one or more 'environmental' measurements.
• This type of experiment is realistic as it is 'natural' and not manipulated. A good general experiment unlike manipulative experiments.
• 2) MANIPULATIVE
• Experimental conditions are manipulated and controlled to investigate cause and effect relationship.
• This approach allows relationships to be deduced.
• Manipulative experiment has excelled control of independent variables and good deduction ability unlike observation experiments.