User Tools

Site Tools


overall_limitations

Overall Limitations

To be able to judge about the evidence delivered by the conducted research it is important to out-line the limits of interpretation of the stated findings (results)

Sample:

As a survey on the total population of the researched regions is not feasible, a sampling took place. To only conduct a research on a sample and than derive conclusions from the sample about the total population is a standard approach within science. However to draw conclusion for the whole popula-tion from a sample it is important to chosen an appropriate and clear sampling methodology. While a not well done sampling technique or not well applied sampling methodology will deform the findings. (Micheel 2010, p. 65-66) A key issue within sampling techniques is the concept of representatives. This concept describes that each person within a population has the same possibility/chance to be chosen as an part of the sample. Representativeness is usually guaranteed by an random sampling technique. (. This methodology don’t lead to a sample which is an exact figure of the total population, but to a an figure, which can be seen as similar ( Nährungsweisesn abbild der grundgesamtheit). So a random sample doesn’t lead to a figure in which the features (age, gender) are proportional to the total population (Heinz-Günter Micheel 2010, p.67-69). If a random sample should be used for drawing conclusion of the distribution of features within the total population it is strongly dependent on the sample size. (ebd.) It can be stated that this sampling technique was applied in the research of the research project “Rural areas in Brandenburg”.

In the Potsdam region another sampling technique was applied. This can be seen as an mixture be-tween an arbitrary sampling technique and an conscious sample. Though first the research project asked arbitrary chosen people within public places, which is a common methodology within student research , caused by limited capacities (finacial) (Micheel 2010, p. 66-67). Further it can be stated that this sampling took place within several places within Potsdam. Moreover, the research group tried to adjusted the sampling method by an conscious selection of participants to get a proportional figure (gender/agee) of the total population. However as this is not a random sampling technique the conclu-sions drawn form the sample are also limited.

Cross section studies (Querschnittsstudie)

As both surveys took place during a short time period they have to be seen as a cross-section studies. Cross-secton studies are common within science and usually chosen through out limited financial re-sources and used for an explorative purpose. These studies can just provide a snap-shot perspective instead of long-time evaluation. In general it can be stated, that this studies are not designed to assess social processes, because social process underlie a specific dynamic. However, it can be stated, that cross-section studies are often applied in social science to conclude about social processes. But it has to be mentioned, that all results delivered by a cross-section analysis are very limited (Micheel 2010, p. 57).

During the survey period a deformation from the results can take place througght out short time trends ( especially media) and external events. During the research period a view media trend and events could have played a role in influencing the answer behaviour of the participants: Media etc. influence

So during the survey period on the national level the EEG-reform was discussed. This can affect the opinion of the locals, because the EEG became a symbol for an increase of electricity prices within the last years. On the federal level of Brandenburg an major event within this time was further the occupation of a coalmine in the south east of Brandenburg, done by the movement “ Ende Gelän-de” 13-16.05.16). This event might not appear to be directly linked to the opinion on wind energy. However, it is linked to an very strong debate around the importance of brown coal mining within Brandenburg and was discussed controversially during the survey period in the media. As there was a lot of disagreement with the type of the actions of the “Ende Gelände” movement. It could be the case, that people identified the “Ende Gelände” Movement as an representative for renewable energy supporters and that the disagreement with the protest action fostered their opposing position towards renewables energy. Further some local events in Uebigau Wahrenbrück could have played a role like the seventh local energy conference, which took place on the third of July 2016. And further there was an participa-tion phase regarding a new spatial plan (Teilflächennutzungsplan) from the 21.04.2016 to the 26.5.2016, in which the locals could among several topics also express their concerns towards the planning of new wind energy development areas. (amtsblatt Nr. 4/2016)- Further event which played a role within the brandenburg area, was a referendum against new windparks, which collected signatures and supporters during that time. ( and planning process of new windfarms in havelland-flämming as well. )

Style of questions:

Chosen question style affect answers of interviewed:

Open questions:

Positiv: possiblitiy for a “uninfluenced” answer Negative: higher requirements for participants, because have to be able to express themselves and formulate their answer on their own. Another problem which goes hand in hand with open questions is that the grouping and coding of answers of open question is done afterwards by the researchers and therefore time consuming (and could be antworten verzerren). However short open question are seen as less problematic. (Micheel 2010, p. 79)

Closed questions & scale questions

Answer possibilities are determined. In contrast to open questions the participants can’t express their detailed opinion about a topic. So closed question leads to an reduced picture of the opinion of the partici-pants – kursiv= steht nicht explizit in der Quelle. As the questionnaires in both research projects contained several questions based on answer scales ( ratings), this issue has to be discussed as well.

Questions involving an scale of answer possibility. The numbers of scale steps is still a subject of con-troversy in literature so some author conclude that a scale range containing 7 or more scale steps , while other recommend 5( Preston/Colmann 2000, Rohrmann 1978 in Micheel 2010, p. 82). However, since the beginning of the 90s a rating scale containing 4 answer possibilities is dominant within sociological research. The question if an even or uneven numbers of scale should be used, the literature recommend, just using an uneven number of a scale, if it is really realistic that a per-son has a “neutral” view on something. Because otherwise a lot of participants tend to chose the neutral option and this can deform the researched opinion within the end. (Micheel 2010, p. 82). In terms of validity and reliability. Measures which tell if the applied methodology measures the right issue and to the right extent. Through out pretest of the designed questionnaires ( excluding the havelland) the methodology was tested. And adapted.

Survey methods

Adding interview limits: Potsdam: Standardisier questionnaire used for interviewer→ (less influence of interviewer than in open interviews) Positive: little interpretation possibility. Questions , in case of not understaning possible. Negativ: Still influences of interviewer. In case of face to face interviewer communication skills as well as appearens ( erscheinungsbild) play a high role. Further in case of special topics the inter-view should fit to the researched group age/gender etc. (Micheel 2010, p. 90-91) Tendency to lie in regards to social acceptance in case of interviewer higher than write question-naires.

-A problem with questionnaire is that respondents may lie due to social desirability. Most people want to present a positive image of themselves and so may lie or bend the truth to look good.

Havelland: written questionnaires / written survey

Written survey: can be delivered by post, mail or other options. Limits: Despite an introduction usualy questionaires includes an description how to answer the questionaire. Return: depending of several influences. Introduction letter usually good for motivating people to respond. Participation in an written survey always depending on the interest of the interviewed. This leads to a biased selection of participants. (So people with a higher interest or a strong view tend to pa-rictipacte more than people with a neutral attitude towrdes the researched topic). However it can be stated that people with a higher education more often showing the willingness to participate in written questionaires. Recommended in literature If there is no responds a reminder letter should be sent ( not than limited fincial and time resources). Further it can be mentioned than within wirt-ten survey the situation cant be seen as totally controllable. So events on the small scale which can influence the answering attitude can appear. (Micheel 2010, p.92)

Overall comparison

Building an index: An index is built by a combination of several variabels. Social acceptance ( dimension) Indicator 1: value wind energy Indicator 2: value renewable Building of the indicator was based on theoretical background. ( however highly reflects opinion of researchers, which question are chosen to assess this issue) If indicatoren can be summarized can be assessed using statistical approaches eg. “konfirma-tiorische Faktoren analyse”. The used index in the comparison is based on weighting of the variables. Based on theoretical thinking, not on statistical assumption. In literature a theoretical reasoning for weighting indicators as problematic and it is argued for statistical processes.( recommended using regressionskoef-fizeint to estimate weightning of variables) (Micheel 2010, p.47-49)

Fusion or merging of samples and data (Baur 2014, p. 367) Huge Sample sizes are very costly. Therefore merging of data is a common practice. . Within a data integration a Variable form different survey are merged. ( data regarding the same individual conducted throughout several surveys can be merged, based on an identification key ) ( data fusion is a Missing-data technique – in the case data for a specific case is missing it can be replaced by an other case which shows statistic similarities)

overall_limitations.txt · Last modified: 2016/07/14 13:37 by leon-thomas