Mafomu eStatistical muKutsvaga
Statistics ibazi remasvomhu rine chekuita nekuunganidza, kuongorora, kududzira, uye kuratidzwa kwedata. Mukutsvaga, kungave mune zvesainzi, mainjiniya, sainzi yemagariro evanhu, kana kunyangwe mune zvehunhu, statistics inoita basa rakakosha mukubatsira vaongorori kuyedza fungidziro, kufanotaura, uye kuwana mhedziso. Chinyorwa chino chichakurukura mamwe mafomura ekutanga ehuwandu hwehuwandu uye mashandisirwo awo mukutsvaga.
1. Nhamba Dzinotsanangura
Nhamba dzinotsanangura dzinoshandiswa kutsanangura data rakaunganidzwa muchidzidzo. Dzinosanganisira zviyero zvakasiyana-siyana zvinopa ruzivo rwese rwedata.
a. Avhareji (Avhareji)
Avhareji ndiyo inonyanya kushandiswa muhuwandu hwe ...
\[ \text{Mean} (\bar{x}) = \frac{\sum_{i=1}^{n} x_i}{n} \]
Di mana:
– \( \sum \) chiratidzo chehuwandu, zvinoreva kubatanidza zvese kukosha kwe \( x \) kubva pa1 kusvika pa \( n \).
– \( x_i \) ndiyo kukosha kwese kuri mudataset.
– \( n \) ndiyo huwandu hwese hwezviyero zviri mudataset.
b. Pakati
Mupakati ndiwo mutengo wepakati uri mudata rakarongeka. Kana nhamba yemitengo iri mudata iri isina kujairika, mupakati ndiwo mutengo wepakati. Kana nhamba yemitengo iri yakaenzana, mupakati ndiwo avhareji yemitengo miviri yepakati.
c. Maitiro
Iyo modhi ndiyo kukosha kunowanzoonekwa museti yedata. Seti yedata inogona kunge iine modhi imwe chete (unimodal), modhi dzinopfuura imwe (multimodal), kana kusava nemodhi zvachose.
d. Nzvimbo
Range ndiyo musiyano uripo pakati pehuwandu hwepamusoro nehwepasi mu dataset.
\[ \text{Range} = \text{Max}(x) – \text{Min}(x) \]
e. Kutsauka Kwakajairwa
Kutsauka kwakajairika inzira yekuyera kupararira kana kupararira kwedata rakatenderedza avhareji. Fomura yekutsauka kwakajairika kwevanhu ndeiyi:
\[ \sigma = \sqrt{\frac{\sum_{i=1}^{N} (x_i – \mu)^2}{N}} \]
Uye yemuenzaniso:
\[ s = \sqrt{\frac{\sum_{i=1}^{n} (x_i – \bar{x})^2}{n-1}} \]
Di mana:
– \( \sigma \) ndiko kutsauka kwehuwandu hwevanhu.
– \( s \) ndiyo muenzaniso wekutsauka kwakajairika.
– \( x_i \) ndiyo kukosha kwese kuri mudataset.
– \( \mu \) ndiyo huwandu hwevanhu.
– \( \bar{x} \) ndiyo muenzaniso wepakati.
– \( N \) ndiyo huwandu hwese hwezviyero zviri muhuwandu hwevanhu.
– \( n \) ndiyo huwandu hwese hwezviyero zviri mumuenzaniso.
2. Nhamba dzekufungidzira
Nhamba dzekufungidzira dzinobvumira vaongorori kuwana mhedziso pamusoro pehuwandu hwevanhu zvichibva pane imwe data. Inosanganisira nzira dzakasiyana-siyana, dzakadai sekuongorora fungidziro, kudzokorora, uye kuongorora kusiyana (ANOVA).
a. Kuedzwa kweFungidziro
Kuongororwa kwefungidziro inzira inoshandiswa kuona kana paine humbowo hwakakwana mumuenzaniso wedata kuti pave nechokwadi chekuti chirwere chiripo muvanhu.
i. Kufungidzira Kusina Kunaka (H0) uye Kufungidzira Kumwe (H1)
– Null Hypothesis (H0): Hapana musiyano kana mhedzisiro.
– Imwe Fungidziro (H1): Pane musiyano kana mhedzisiro.
Kuedza kunoitwa uchishandisa nhamba yekuedza, senge t-test, chi-square test, kana ANOVA, uye kuenzanisa p-value ne significance level (\(\alpha\)), kazhinji 0,05.
ii. bvunzo ye t
Bvunzo reT rinoshandiswa kuenzanisa nzira dzemapoka maviri. Kune misiyano yakati wandei yebvunzo reT, senge sampuli dzakazvimiririra dzeT-test neT-test dzakapairwa.
Fomura yekutanga ye t-test yemasampuli akazvimirira ndeiyi:
\[ t = \frac{\bar{x}_1 – \bar{x}_2}{\sqrt{\left( \frac{s_1^2}{n_1} \right) + \left( \frac{s_2^2}{n_2} \right)}} \]
b. Kudzokera shure
Kudzoreredza kunoshandiswa kuratidza hukama huripo pakati peimwe kana kupfuura mavariable akazvimiririra (predictors) uye dependent variable (response).
i. Kudzoreredzwa Kwemutsetse Kuri Nyore
Kudzokorora kuri nyore kunoratidza hukama huripo pakati pechinhu chimwe chete chakazvimiririra uye chimwe chete chinoenderana.
Equation iri nyore yekudzokorora mutsara ndeiyi:
\[ y = \beta_0 + \beta_1 x + \epsilon \]
Di mana:
– \( y \) ishanduro inoenderana.
– \( x \) ishanduro yakazvimiririra.
– \( \beta_0 \) ndiyo intercept.
– \( \beta_1 \) ndiyo regression coefficient.
– \( \epsilon \) chikanganiso.
ii. Kudzoreredzwa Kwemitsara Yakawanda
Mamodheru ekudzokorora mutsara akawanda anoratidza hukama huripo pakati pemavariable akawanda akazvimiririra uye imwe dependent variable.
Equation ye multiple linear regression ndeiyi:
\[ y = \beta_0 + \beta_1 x_1 + \beta_2 x_2 + \ldots + \beta_p x_p + \epsilon \]
Di mana:
– \( y \) ishanduro inoenderana.
– \( x_1, x_2, \ldots, x_p \) ndiyo variable yakazvimiririra.
– \( \beta_0 \) ndiyo intercept.
– \( \beta_p \) ndiyo regression coefficient ye variable yakazvimirira \( p \).
– \( \epsilon \) chikanganiso.
c. Kuongororwa kweKusiyana (ANOVA)
ANOVA inoshandiswa kuenzanisa nzira dzemapoka matatu kana kupfuura. ANOVA inoona kana paine musiyano mukuru pakati penzira dzemapoka nekuenzanisa musiyano uripo pakati pemapoka nekusiyana kuripo mukati memapoka.
Fomura yekutanga yeANOVA ndeiyi:
\[ F = \frac{\text{Between-Group Variability}}{\text{Within-Group Variability}} \]
Nhamba yeF inoverengerwa uye inoenzaniswa nehukuru hwekugoverwa kweF kuti ione kana paine musiyano mukuru pakati pehuwandu hweboka.
3. Kubatana
Kubatana kunoyera simba negwara rehukama hwakatsetseka pakati pezvinhu zviviri.
a. Pearson Correlation Coefficient (r)
Chiyero chePearson chekubatana ndicho chinoshandiswa zvakanyanya kuyera hukama hwakatsetseka pakati pezvinhu zviviri.
Fomura yePearson coefficient ndeye:
\[ r = \frac{\sum_{i=1}^{n} (x_i – \bar{x})(y_i – \bar{y})}{\sqrt{\sum_{i=1}^{n} (x_i – \bar{x})^2} \sqrt{\sum_{i=1}^{n} (y_i – \bar{y})^2}} \]
Di mana:
– \( r \) ndiyo coefficient yekubatana kwePearson.
– \( x_i \) uye \( y_i \) ndiwo mavalues ezvimiro zviviri.
– \( \bar{x} \) uye \( \bar{y} \) ndiwo maavhareji ezvimiro zviviri izvi.
Kukosha kwe \( r \) kunotangira pa -1 (perfect negative correlation) kusvika pa +1 (perfect positive correlation), ne 0 ichiratidza kuti hapana correlation.
Penutup
Nhamba dzezviverengero chishandiso chakakosha chekutsvagisa chinobatsira vaongorori kupa, kuongorora, uye kuwana mhedziso kubva padata. Chinyorwa chino chinongotaura nezvemafomula mashoma ekutanga mutsananguro nenhamba dzekufungidzira, pamwe nekubatana. Kunyange zvazvo zviri nyore, kunzwisisa kwakakwana kwemafomula aya kwakakosha pakuita ongororo dzakakodzera uye kuwana mhedziso dzakarurama kubva padata rekutsvagisa. Nekuziva nhamba, vaongorori vanogona kuve nechokwadi chekuti zvavakawana zvakavakirwa pakuongorora kwakasimba uye kwakavimbika.
